diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,31172 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999155096178218,
+  "eval_steps": 500,
+  "global_step": 4438,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00022530768580843213,
+      "grad_norm": 1.0947727710467006,
+      "learning_rate": 2.2522522522522524e-08,
+      "loss": 0.3315,
+      "step": 1
+    },
+    {
+      "epoch": 0.00045061537161686426,
+      "grad_norm": 1.1056270284189085,
+      "learning_rate": 4.504504504504505e-08,
+      "loss": 0.3365,
+      "step": 2
+    },
+    {
+      "epoch": 0.0006759230574252964,
+      "grad_norm": 1.1038791662741807,
+      "learning_rate": 6.756756756756757e-08,
+      "loss": 0.3123,
+      "step": 3
+    },
+    {
+      "epoch": 0.0009012307432337285,
+      "grad_norm": 1.1319684508304422,
+      "learning_rate": 9.00900900900901e-08,
+      "loss": 0.3496,
+      "step": 4
+    },
+    {
+      "epoch": 0.0011265384290421608,
+      "grad_norm": 1.0950409822450369,
+      "learning_rate": 1.1261261261261262e-07,
+      "loss": 0.3356,
+      "step": 5
+    },
+    {
+      "epoch": 0.0013518461148505929,
+      "grad_norm": 1.110496283834593,
+      "learning_rate": 1.3513513513513515e-07,
+      "loss": 0.3396,
+      "step": 6
+    },
+    {
+      "epoch": 0.001577153800659025,
+      "grad_norm": 1.0802753121404411,
+      "learning_rate": 1.5765765765765766e-07,
+      "loss": 0.3135,
+      "step": 7
+    },
+    {
+      "epoch": 0.001802461486467457,
+      "grad_norm": 1.1166340984108152,
+      "learning_rate": 1.801801801801802e-07,
+      "loss": 0.3234,
+      "step": 8
+    },
+    {
+      "epoch": 0.0020277691722758893,
+      "grad_norm": 1.0777930663732238,
+      "learning_rate": 2.0270270270270273e-07,
+      "loss": 0.3248,
+      "step": 9
+    },
+    {
+      "epoch": 0.0022530768580843216,
+      "grad_norm": 1.0647347889926293,
+      "learning_rate": 2.2522522522522524e-07,
+      "loss": 0.3323,
+      "step": 10
+    },
+    {
+      "epoch": 0.0024783845438927535,
+      "grad_norm": 1.0702813501737287,
+      "learning_rate": 2.477477477477478e-07,
+      "loss": 0.3538,
+      "step": 11
+    },
+    {
+      "epoch": 0.0027036922297011858,
+      "grad_norm": 1.0776989359324258,
+      "learning_rate": 2.702702702702703e-07,
+      "loss": 0.3381,
+      "step": 12
+    },
+    {
+      "epoch": 0.0029289999155096176,
+      "grad_norm": 1.0547392159345734,
+      "learning_rate": 2.927927927927928e-07,
+      "loss": 0.325,
+      "step": 13
+    },
+    {
+      "epoch": 0.00315430760131805,
+      "grad_norm": 1.1128330063858483,
+      "learning_rate": 3.153153153153153e-07,
+      "loss": 0.3457,
+      "step": 14
+    },
+    {
+      "epoch": 0.003379615287126482,
+      "grad_norm": 1.026561868913419,
+      "learning_rate": 3.378378378378379e-07,
+      "loss": 0.3302,
+      "step": 15
+    },
+    {
+      "epoch": 0.003604922972934914,
+      "grad_norm": 1.0844063716667791,
+      "learning_rate": 3.603603603603604e-07,
+      "loss": 0.3511,
+      "step": 16
+    },
+    {
+      "epoch": 0.0038302306587433463,
+      "grad_norm": 1.0432870057572243,
+      "learning_rate": 3.828828828828829e-07,
+      "loss": 0.3313,
+      "step": 17
+    },
+    {
+      "epoch": 0.004055538344551779,
+      "grad_norm": 1.0877367387628776,
+      "learning_rate": 4.0540540540540546e-07,
+      "loss": 0.3554,
+      "step": 18
+    },
+    {
+      "epoch": 0.0042808460303602105,
+      "grad_norm": 1.055171840020527,
+      "learning_rate": 4.27927927927928e-07,
+      "loss": 0.3501,
+      "step": 19
+    },
+    {
+      "epoch": 0.004506153716168643,
+      "grad_norm": 1.0226585393284515,
+      "learning_rate": 4.504504504504505e-07,
+      "loss": 0.3312,
+      "step": 20
+    },
+    {
+      "epoch": 0.004731461401977075,
+      "grad_norm": 0.9662290728315617,
+      "learning_rate": 4.7297297297297305e-07,
+      "loss": 0.3339,
+      "step": 21
+    },
+    {
+      "epoch": 0.004956769087785507,
+      "grad_norm": 0.9497175074415394,
+      "learning_rate": 4.954954954954956e-07,
+      "loss": 0.3321,
+      "step": 22
+    },
+    {
+      "epoch": 0.00518207677359394,
+      "grad_norm": 0.9901701348569355,
+      "learning_rate": 5.180180180180181e-07,
+      "loss": 0.3265,
+      "step": 23
+    },
+    {
+      "epoch": 0.0054073844594023715,
+      "grad_norm": 0.8828316210132122,
+      "learning_rate": 5.405405405405406e-07,
+      "loss": 0.3109,
+      "step": 24
+    },
+    {
+      "epoch": 0.005632692145210803,
+      "grad_norm": 0.906520614071178,
+      "learning_rate": 5.630630630630631e-07,
+      "loss": 0.3252,
+      "step": 25
+    },
+    {
+      "epoch": 0.005857999831019235,
+      "grad_norm": 0.8837738853128801,
+      "learning_rate": 5.855855855855856e-07,
+      "loss": 0.3313,
+      "step": 26
+    },
+    {
+      "epoch": 0.006083307516827668,
+      "grad_norm": 0.8839821083094102,
+      "learning_rate": 6.081081081081082e-07,
+      "loss": 0.3475,
+      "step": 27
+    },
+    {
+      "epoch": 0.0063086152026361,
+      "grad_norm": 0.8092230130785764,
+      "learning_rate": 6.306306306306306e-07,
+      "loss": 0.3134,
+      "step": 28
+    },
+    {
+      "epoch": 0.006533922888444532,
+      "grad_norm": 0.7697471952605461,
+      "learning_rate": 6.531531531531532e-07,
+      "loss": 0.3143,
+      "step": 29
+    },
+    {
+      "epoch": 0.006759230574252964,
+      "grad_norm": 0.7360077029834586,
+      "learning_rate": 6.756756756756758e-07,
+      "loss": 0.2975,
+      "step": 30
+    },
+    {
+      "epoch": 0.006984538260061396,
+      "grad_norm": 0.7303882904067905,
+      "learning_rate": 6.981981981981982e-07,
+      "loss": 0.3138,
+      "step": 31
+    },
+    {
+      "epoch": 0.007209845945869828,
+      "grad_norm": 0.7331410718564746,
+      "learning_rate": 7.207207207207208e-07,
+      "loss": 0.3118,
+      "step": 32
+    },
+    {
+      "epoch": 0.007435153631678261,
+      "grad_norm": 0.7585176643664102,
+      "learning_rate": 7.432432432432434e-07,
+      "loss": 0.3238,
+      "step": 33
+    },
+    {
+      "epoch": 0.007660461317486693,
+      "grad_norm": 0.6929989393254152,
+      "learning_rate": 7.657657657657658e-07,
+      "loss": 0.2965,
+      "step": 34
+    },
+    {
+      "epoch": 0.007885769003295125,
+      "grad_norm": 0.7248401729174301,
+      "learning_rate": 7.882882882882883e-07,
+      "loss": 0.3134,
+      "step": 35
+    },
+    {
+      "epoch": 0.008111076689103557,
+      "grad_norm": 0.6846514226447453,
+      "learning_rate": 8.108108108108109e-07,
+      "loss": 0.2876,
+      "step": 36
+    },
+    {
+      "epoch": 0.008336384374911989,
+      "grad_norm": 0.6848701434990109,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 0.2921,
+      "step": 37
+    },
+    {
+      "epoch": 0.008561692060720421,
+      "grad_norm": 0.684924752449004,
+      "learning_rate": 8.55855855855856e-07,
+      "loss": 0.3097,
+      "step": 38
+    },
+    {
+      "epoch": 0.008786999746528853,
+      "grad_norm": 0.6598684646273593,
+      "learning_rate": 8.783783783783785e-07,
+      "loss": 0.2996,
+      "step": 39
+    },
+    {
+      "epoch": 0.009012307432337286,
+      "grad_norm": 0.651883046062235,
+      "learning_rate": 9.00900900900901e-07,
+      "loss": 0.2918,
+      "step": 40
+    },
+    {
+      "epoch": 0.009237615118145718,
+      "grad_norm": 0.6607351963566211,
+      "learning_rate": 9.234234234234235e-07,
+      "loss": 0.3159,
+      "step": 41
+    },
+    {
+      "epoch": 0.00946292280395415,
+      "grad_norm": 0.5841038932200996,
+      "learning_rate": 9.459459459459461e-07,
+      "loss": 0.2989,
+      "step": 42
+    },
+    {
+      "epoch": 0.009688230489762582,
+      "grad_norm": 0.5958613029441983,
+      "learning_rate": 9.684684684684686e-07,
+      "loss": 0.2812,
+      "step": 43
+    },
+    {
+      "epoch": 0.009913538175571014,
+      "grad_norm": 0.6099154207191207,
+      "learning_rate": 9.909909909909911e-07,
+      "loss": 0.2793,
+      "step": 44
+    },
+    {
+      "epoch": 0.010138845861379446,
+      "grad_norm": 0.6118645853590743,
+      "learning_rate": 1.0135135135135136e-06,
+      "loss": 0.2754,
+      "step": 45
+    },
+    {
+      "epoch": 0.01036415354718788,
+      "grad_norm": 0.6542805689870427,
+      "learning_rate": 1.0360360360360361e-06,
+      "loss": 0.2822,
+      "step": 46
+    },
+    {
+      "epoch": 0.010589461232996311,
+      "grad_norm": 0.6335055926076487,
+      "learning_rate": 1.0585585585585587e-06,
+      "loss": 0.2655,
+      "step": 47
+    },
+    {
+      "epoch": 0.010814768918804743,
+      "grad_norm": 0.6801605196071887,
+      "learning_rate": 1.0810810810810812e-06,
+      "loss": 0.2919,
+      "step": 48
+    },
+    {
+      "epoch": 0.011040076604613175,
+      "grad_norm": 0.5974596009318522,
+      "learning_rate": 1.1036036036036037e-06,
+      "loss": 0.2799,
+      "step": 49
+    },
+    {
+      "epoch": 0.011265384290421607,
+      "grad_norm": 0.6125372348195963,
+      "learning_rate": 1.1261261261261262e-06,
+      "loss": 0.2759,
+      "step": 50
+    },
+    {
+      "epoch": 0.011490691976230039,
+      "grad_norm": 0.5838366991561806,
+      "learning_rate": 1.148648648648649e-06,
+      "loss": 0.278,
+      "step": 51
+    },
+    {
+      "epoch": 0.01171599966203847,
+      "grad_norm": 0.6341141792855924,
+      "learning_rate": 1.1711711711711712e-06,
+      "loss": 0.2865,
+      "step": 52
+    },
+    {
+      "epoch": 0.011941307347846904,
+      "grad_norm": 0.615509564485624,
+      "learning_rate": 1.1936936936936937e-06,
+      "loss": 0.2848,
+      "step": 53
+    },
+    {
+      "epoch": 0.012166615033655336,
+      "grad_norm": 0.5416726529744037,
+      "learning_rate": 1.2162162162162164e-06,
+      "loss": 0.2662,
+      "step": 54
+    },
+    {
+      "epoch": 0.012391922719463768,
+      "grad_norm": 0.5686478139322753,
+      "learning_rate": 1.2387387387387387e-06,
+      "loss": 0.3033,
+      "step": 55
+    },
+    {
+      "epoch": 0.0126172304052722,
+      "grad_norm": 0.5706950595773004,
+      "learning_rate": 1.2612612612612613e-06,
+      "loss": 0.2808,
+      "step": 56
+    },
+    {
+      "epoch": 0.012842538091080631,
+      "grad_norm": 0.5342854152172102,
+      "learning_rate": 1.2837837837837838e-06,
+      "loss": 0.2682,
+      "step": 57
+    },
+    {
+      "epoch": 0.013067845776889063,
+      "grad_norm": 0.5320452417181604,
+      "learning_rate": 1.3063063063063065e-06,
+      "loss": 0.2528,
+      "step": 58
+    },
+    {
+      "epoch": 0.013293153462697497,
+      "grad_norm": 0.5435487202598747,
+      "learning_rate": 1.328828828828829e-06,
+      "loss": 0.273,
+      "step": 59
+    },
+    {
+      "epoch": 0.013518461148505929,
+      "grad_norm": 0.5413964558339608,
+      "learning_rate": 1.3513513513513515e-06,
+      "loss": 0.2731,
+      "step": 60
+    },
+    {
+      "epoch": 0.01374376883431436,
+      "grad_norm": 0.5348770326410263,
+      "learning_rate": 1.373873873873874e-06,
+      "loss": 0.2749,
+      "step": 61
+    },
+    {
+      "epoch": 0.013969076520122792,
+      "grad_norm": 0.5318413005939099,
+      "learning_rate": 1.3963963963963963e-06,
+      "loss": 0.2672,
+      "step": 62
+    },
+    {
+      "epoch": 0.014194384205931224,
+      "grad_norm": 0.4899105874773047,
+      "learning_rate": 1.418918918918919e-06,
+      "loss": 0.2403,
+      "step": 63
+    },
+    {
+      "epoch": 0.014419691891739656,
+      "grad_norm": 0.5545085049603222,
+      "learning_rate": 1.4414414414414416e-06,
+      "loss": 0.2961,
+      "step": 64
+    },
+    {
+      "epoch": 0.01464499957754809,
+      "grad_norm": 0.529025173817156,
+      "learning_rate": 1.463963963963964e-06,
+      "loss": 0.2717,
+      "step": 65
+    },
+    {
+      "epoch": 0.014870307263356522,
+      "grad_norm": 0.5550818812254924,
+      "learning_rate": 1.4864864864864868e-06,
+      "loss": 0.2742,
+      "step": 66
+    },
+    {
+      "epoch": 0.015095614949164953,
+      "grad_norm": 0.5398535606143846,
+      "learning_rate": 1.5090090090090093e-06,
+      "loss": 0.2688,
+      "step": 67
+    },
+    {
+      "epoch": 0.015320922634973385,
+      "grad_norm": 0.5106467899782415,
+      "learning_rate": 1.5315315315315316e-06,
+      "loss": 0.2476,
+      "step": 68
+    },
+    {
+      "epoch": 0.015546230320781817,
+      "grad_norm": 0.5333505346961758,
+      "learning_rate": 1.5540540540540541e-06,
+      "loss": 0.2584,
+      "step": 69
+    },
+    {
+      "epoch": 0.01577153800659025,
+      "grad_norm": 0.5256415895130913,
+      "learning_rate": 1.5765765765765766e-06,
+      "loss": 0.2711,
+      "step": 70
+    },
+    {
+      "epoch": 0.015996845692398683,
+      "grad_norm": 0.5184587920114005,
+      "learning_rate": 1.5990990990990993e-06,
+      "loss": 0.2768,
+      "step": 71
+    },
+    {
+      "epoch": 0.016222153378207115,
+      "grad_norm": 0.5163158029073385,
+      "learning_rate": 1.6216216216216219e-06,
+      "loss": 0.2701,
+      "step": 72
+    },
+    {
+      "epoch": 0.016447461064015546,
+      "grad_norm": 0.48032723537669814,
+      "learning_rate": 1.6441441441441444e-06,
+      "loss": 0.2344,
+      "step": 73
+    },
+    {
+      "epoch": 0.016672768749823978,
+      "grad_norm": 0.48140088563311695,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.2498,
+      "step": 74
+    },
+    {
+      "epoch": 0.01689807643563241,
+      "grad_norm": 0.49290629978075023,
+      "learning_rate": 1.6891891891891894e-06,
+      "loss": 0.2544,
+      "step": 75
+    },
+    {
+      "epoch": 0.017123384121440842,
+      "grad_norm": 0.4967869263831517,
+      "learning_rate": 1.711711711711712e-06,
+      "loss": 0.2417,
+      "step": 76
+    },
+    {
+      "epoch": 0.017348691807249274,
+      "grad_norm": 0.5096265115174715,
+      "learning_rate": 1.7342342342342344e-06,
+      "loss": 0.2439,
+      "step": 77
+    },
+    {
+      "epoch": 0.017573999493057706,
+      "grad_norm": 0.5302741235480698,
+      "learning_rate": 1.756756756756757e-06,
+      "loss": 0.2641,
+      "step": 78
+    },
+    {
+      "epoch": 0.017799307178866138,
+      "grad_norm": 0.4832401096007166,
+      "learning_rate": 1.7792792792792792e-06,
+      "loss": 0.2385,
+      "step": 79
+    },
+    {
+      "epoch": 0.018024614864674573,
+      "grad_norm": 0.5169432406664927,
+      "learning_rate": 1.801801801801802e-06,
+      "loss": 0.2616,
+      "step": 80
+    },
+    {
+      "epoch": 0.018249922550483005,
+      "grad_norm": 0.4810264047446268,
+      "learning_rate": 1.8243243243243245e-06,
+      "loss": 0.235,
+      "step": 81
+    },
+    {
+      "epoch": 0.018475230236291437,
+      "grad_norm": 0.5064455552910841,
+      "learning_rate": 1.846846846846847e-06,
+      "loss": 0.2502,
+      "step": 82
+    },
+    {
+      "epoch": 0.01870053792209987,
+      "grad_norm": 0.5252312972000183,
+      "learning_rate": 1.8693693693693697e-06,
+      "loss": 0.2527,
+      "step": 83
+    },
+    {
+      "epoch": 0.0189258456079083,
+      "grad_norm": 0.5152115480424759,
+      "learning_rate": 1.8918918918918922e-06,
+      "loss": 0.252,
+      "step": 84
+    },
+    {
+      "epoch": 0.019151153293716732,
+      "grad_norm": 0.4792935284601226,
+      "learning_rate": 1.9144144144144145e-06,
+      "loss": 0.2239,
+      "step": 85
+    },
+    {
+      "epoch": 0.019376460979525164,
+      "grad_norm": 0.46930021033135166,
+      "learning_rate": 1.9369369369369372e-06,
+      "loss": 0.226,
+      "step": 86
+    },
+    {
+      "epoch": 0.019601768665333596,
+      "grad_norm": 0.5119347999323998,
+      "learning_rate": 1.9594594594594595e-06,
+      "loss": 0.2479,
+      "step": 87
+    },
+    {
+      "epoch": 0.019827076351142028,
+      "grad_norm": 0.5168060113294949,
+      "learning_rate": 1.9819819819819822e-06,
+      "loss": 0.2448,
+      "step": 88
+    },
+    {
+      "epoch": 0.02005238403695046,
+      "grad_norm": 0.5163631879799307,
+      "learning_rate": 2.0045045045045045e-06,
+      "loss": 0.2382,
+      "step": 89
+    },
+    {
+      "epoch": 0.02027769172275889,
+      "grad_norm": 0.5159666924757886,
+      "learning_rate": 2.0270270270270273e-06,
+      "loss": 0.2418,
+      "step": 90
+    },
+    {
+      "epoch": 0.020502999408567323,
+      "grad_norm": 0.5041283221176179,
+      "learning_rate": 2.0495495495495496e-06,
+      "loss": 0.2659,
+      "step": 91
+    },
+    {
+      "epoch": 0.02072830709437576,
+      "grad_norm": 0.5016352886414731,
+      "learning_rate": 2.0720720720720723e-06,
+      "loss": 0.2473,
+      "step": 92
+    },
+    {
+      "epoch": 0.02095361478018419,
+      "grad_norm": 0.4805339701948201,
+      "learning_rate": 2.0945945945945946e-06,
+      "loss": 0.2394,
+      "step": 93
+    },
+    {
+      "epoch": 0.021178922465992622,
+      "grad_norm": 0.5063248501917516,
+      "learning_rate": 2.1171171171171173e-06,
+      "loss": 0.2531,
+      "step": 94
+    },
+    {
+      "epoch": 0.021404230151801054,
+      "grad_norm": 0.5392779967753892,
+      "learning_rate": 2.13963963963964e-06,
+      "loss": 0.2303,
+      "step": 95
+    },
+    {
+      "epoch": 0.021629537837609486,
+      "grad_norm": 0.5231578606978835,
+      "learning_rate": 2.1621621621621623e-06,
+      "loss": 0.2611,
+      "step": 96
+    },
+    {
+      "epoch": 0.021854845523417918,
+      "grad_norm": 0.49267022849605074,
+      "learning_rate": 2.1846846846846846e-06,
+      "loss": 0.241,
+      "step": 97
+    },
+    {
+      "epoch": 0.02208015320922635,
+      "grad_norm": 0.5362283704864,
+      "learning_rate": 2.2072072072072073e-06,
+      "loss": 0.2622,
+      "step": 98
+    },
+    {
+      "epoch": 0.02230546089503478,
+      "grad_norm": 0.5223976037466284,
+      "learning_rate": 2.22972972972973e-06,
+      "loss": 0.2646,
+      "step": 99
+    },
+    {
+      "epoch": 0.022530768580843213,
+      "grad_norm": 0.4884083903659108,
+      "learning_rate": 2.2522522522522524e-06,
+      "loss": 0.2343,
+      "step": 100
+    },
+    {
+      "epoch": 0.022756076266651645,
+      "grad_norm": 0.5131811981346736,
+      "learning_rate": 2.274774774774775e-06,
+      "loss": 0.2254,
+      "step": 101
+    },
+    {
+      "epoch": 0.022981383952460077,
+      "grad_norm": 0.563620994127029,
+      "learning_rate": 2.297297297297298e-06,
+      "loss": 0.2339,
+      "step": 102
+    },
+    {
+      "epoch": 0.02320669163826851,
+      "grad_norm": 0.4985639954502683,
+      "learning_rate": 2.31981981981982e-06,
+      "loss": 0.2394,
+      "step": 103
+    },
+    {
+      "epoch": 0.02343199932407694,
+      "grad_norm": 0.49335868238483493,
+      "learning_rate": 2.3423423423423424e-06,
+      "loss": 0.2311,
+      "step": 104
+    },
+    {
+      "epoch": 0.023657307009885376,
+      "grad_norm": 0.5574006983841419,
+      "learning_rate": 2.364864864864865e-06,
+      "loss": 0.2573,
+      "step": 105
+    },
+    {
+      "epoch": 0.023882614695693808,
+      "grad_norm": 0.5573967154584135,
+      "learning_rate": 2.3873873873873874e-06,
+      "loss": 0.2409,
+      "step": 106
+    },
+    {
+      "epoch": 0.02410792238150224,
+      "grad_norm": 0.5318573843929276,
+      "learning_rate": 2.40990990990991e-06,
+      "loss": 0.2308,
+      "step": 107
+    },
+    {
+      "epoch": 0.024333230067310672,
+      "grad_norm": 0.5018813770253644,
+      "learning_rate": 2.432432432432433e-06,
+      "loss": 0.2275,
+      "step": 108
+    },
+    {
+      "epoch": 0.024558537753119104,
+      "grad_norm": 0.4919658976142004,
+      "learning_rate": 2.454954954954955e-06,
+      "loss": 0.2452,
+      "step": 109
+    },
+    {
+      "epoch": 0.024783845438927535,
+      "grad_norm": 0.5172596284253529,
+      "learning_rate": 2.4774774774774775e-06,
+      "loss": 0.219,
+      "step": 110
+    },
+    {
+      "epoch": 0.025009153124735967,
+      "grad_norm": 0.5058848987637898,
+      "learning_rate": 2.5e-06,
+      "loss": 0.2401,
+      "step": 111
+    },
+    {
+      "epoch": 0.0252344608105444,
+      "grad_norm": 0.5146673047789411,
+      "learning_rate": 2.5225225225225225e-06,
+      "loss": 0.2626,
+      "step": 112
+    },
+    {
+      "epoch": 0.02545976849635283,
+      "grad_norm": 0.4705301441723405,
+      "learning_rate": 2.5450450450450452e-06,
+      "loss": 0.214,
+      "step": 113
+    },
+    {
+      "epoch": 0.025685076182161263,
+      "grad_norm": 0.524408956720651,
+      "learning_rate": 2.5675675675675675e-06,
+      "loss": 0.2327,
+      "step": 114
+    },
+    {
+      "epoch": 0.025910383867969695,
+      "grad_norm": 0.5300256351060787,
+      "learning_rate": 2.5900900900900907e-06,
+      "loss": 0.2481,
+      "step": 115
+    },
+    {
+      "epoch": 0.026135691553778127,
+      "grad_norm": 0.5010212711826415,
+      "learning_rate": 2.612612612612613e-06,
+      "loss": 0.234,
+      "step": 116
+    },
+    {
+      "epoch": 0.026360999239586562,
+      "grad_norm": 0.50692577392399,
+      "learning_rate": 2.6351351351351353e-06,
+      "loss": 0.2341,
+      "step": 117
+    },
+    {
+      "epoch": 0.026586306925394994,
+      "grad_norm": 0.5013472652726951,
+      "learning_rate": 2.657657657657658e-06,
+      "loss": 0.2334,
+      "step": 118
+    },
+    {
+      "epoch": 0.026811614611203426,
+      "grad_norm": 0.49763653854948314,
+      "learning_rate": 2.6801801801801803e-06,
+      "loss": 0.2334,
+      "step": 119
+    },
+    {
+      "epoch": 0.027036922297011858,
+      "grad_norm": 0.5058647611858305,
+      "learning_rate": 2.702702702702703e-06,
+      "loss": 0.2319,
+      "step": 120
+    },
+    {
+      "epoch": 0.02726222998282029,
+      "grad_norm": 0.5220281062730923,
+      "learning_rate": 2.7252252252252253e-06,
+      "loss": 0.2258,
+      "step": 121
+    },
+    {
+      "epoch": 0.02748753766862872,
+      "grad_norm": 0.4902295196707689,
+      "learning_rate": 2.747747747747748e-06,
+      "loss": 0.234,
+      "step": 122
+    },
+    {
+      "epoch": 0.027712845354437153,
+      "grad_norm": 0.49334228839146704,
+      "learning_rate": 2.7702702702702703e-06,
+      "loss": 0.2222,
+      "step": 123
+    },
+    {
+      "epoch": 0.027938153040245585,
+      "grad_norm": 0.5140530380734651,
+      "learning_rate": 2.7927927927927926e-06,
+      "loss": 0.246,
+      "step": 124
+    },
+    {
+      "epoch": 0.028163460726054017,
+      "grad_norm": 0.5054001798341684,
+      "learning_rate": 2.8153153153153158e-06,
+      "loss": 0.2261,
+      "step": 125
+    },
+    {
+      "epoch": 0.02838876841186245,
+      "grad_norm": 0.5025232523144892,
+      "learning_rate": 2.837837837837838e-06,
+      "loss": 0.2444,
+      "step": 126
+    },
+    {
+      "epoch": 0.02861407609767088,
+      "grad_norm": 0.501416740003468,
+      "learning_rate": 2.860360360360361e-06,
+      "loss": 0.2448,
+      "step": 127
+    },
+    {
+      "epoch": 0.028839383783479312,
+      "grad_norm": 0.5001885864063282,
+      "learning_rate": 2.882882882882883e-06,
+      "loss": 0.221,
+      "step": 128
+    },
+    {
+      "epoch": 0.029064691469287748,
+      "grad_norm": 0.5058916094632965,
+      "learning_rate": 2.9054054054054054e-06,
+      "loss": 0.2372,
+      "step": 129
+    },
+    {
+      "epoch": 0.02928999915509618,
+      "grad_norm": 0.49414108281771374,
+      "learning_rate": 2.927927927927928e-06,
+      "loss": 0.2449,
+      "step": 130
+    },
+    {
+      "epoch": 0.02951530684090461,
+      "grad_norm": 0.5274282549164447,
+      "learning_rate": 2.9504504504504504e-06,
+      "loss": 0.2204,
+      "step": 131
+    },
+    {
+      "epoch": 0.029740614526713043,
+      "grad_norm": 0.5059068511500548,
+      "learning_rate": 2.9729729729729736e-06,
+      "loss": 0.239,
+      "step": 132
+    },
+    {
+      "epoch": 0.029965922212521475,
+      "grad_norm": 0.5403324542543835,
+      "learning_rate": 2.995495495495496e-06,
+      "loss": 0.2427,
+      "step": 133
+    },
+    {
+      "epoch": 0.030191229898329907,
+      "grad_norm": 0.5129734450730943,
+      "learning_rate": 3.0180180180180186e-06,
+      "loss": 0.2233,
+      "step": 134
+    },
+    {
+      "epoch": 0.03041653758413834,
+      "grad_norm": 0.4995498012664299,
+      "learning_rate": 3.040540540540541e-06,
+      "loss": 0.2384,
+      "step": 135
+    },
+    {
+      "epoch": 0.03064184526994677,
+      "grad_norm": 0.4705902432670821,
+      "learning_rate": 3.063063063063063e-06,
+      "loss": 0.2293,
+      "step": 136
+    },
+    {
+      "epoch": 0.030867152955755203,
+      "grad_norm": 0.47864446491382573,
+      "learning_rate": 3.085585585585586e-06,
+      "loss": 0.2283,
+      "step": 137
+    },
+    {
+      "epoch": 0.031092460641563634,
+      "grad_norm": 0.5237766445850344,
+      "learning_rate": 3.1081081081081082e-06,
+      "loss": 0.2352,
+      "step": 138
+    },
+    {
+      "epoch": 0.03131776832737207,
+      "grad_norm": 0.47357640320589656,
+      "learning_rate": 3.130630630630631e-06,
+      "loss": 0.2305,
+      "step": 139
+    },
+    {
+      "epoch": 0.0315430760131805,
+      "grad_norm": 0.47748642005092856,
+      "learning_rate": 3.1531531531531532e-06,
+      "loss": 0.2173,
+      "step": 140
+    },
+    {
+      "epoch": 0.03176838369898893,
+      "grad_norm": 0.5076332367130231,
+      "learning_rate": 3.1756756756756755e-06,
+      "loss": 0.2211,
+      "step": 141
+    },
+    {
+      "epoch": 0.031993691384797365,
+      "grad_norm": 0.49914085991583146,
+      "learning_rate": 3.1981981981981987e-06,
+      "loss": 0.229,
+      "step": 142
+    },
+    {
+      "epoch": 0.0322189990706058,
+      "grad_norm": 0.5156689222761607,
+      "learning_rate": 3.220720720720721e-06,
+      "loss": 0.247,
+      "step": 143
+    },
+    {
+      "epoch": 0.03244430675641423,
+      "grad_norm": 0.5306167064232364,
+      "learning_rate": 3.2432432432432437e-06,
+      "loss": 0.2321,
+      "step": 144
+    },
+    {
+      "epoch": 0.03266961444222266,
+      "grad_norm": 0.5006105302460152,
+      "learning_rate": 3.265765765765766e-06,
+      "loss": 0.2412,
+      "step": 145
+    },
+    {
+      "epoch": 0.03289492212803109,
+      "grad_norm": 0.47660844123587387,
+      "learning_rate": 3.2882882882882887e-06,
+      "loss": 0.2112,
+      "step": 146
+    },
+    {
+      "epoch": 0.033120229813839525,
+      "grad_norm": 0.4908936026508093,
+      "learning_rate": 3.310810810810811e-06,
+      "loss": 0.2146,
+      "step": 147
+    },
+    {
+      "epoch": 0.033345537499647956,
+      "grad_norm": 0.4917283785130264,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.2115,
+      "step": 148
+    },
+    {
+      "epoch": 0.03357084518545639,
+      "grad_norm": 0.49052234204011164,
+      "learning_rate": 3.3558558558558565e-06,
+      "loss": 0.2091,
+      "step": 149
+    },
+    {
+      "epoch": 0.03379615287126482,
+      "grad_norm": 0.5035124349420905,
+      "learning_rate": 3.3783783783783788e-06,
+      "loss": 0.2232,
+      "step": 150
+    },
+    {
+      "epoch": 0.03402146055707325,
+      "grad_norm": 0.5079586810803105,
+      "learning_rate": 3.4009009009009015e-06,
+      "loss": 0.2283,
+      "step": 151
+    },
+    {
+      "epoch": 0.034246768242881684,
+      "grad_norm": 0.5022741231517249,
+      "learning_rate": 3.423423423423424e-06,
+      "loss": 0.2043,
+      "step": 152
+    },
+    {
+      "epoch": 0.034472075928690116,
+      "grad_norm": 0.49004074919119595,
+      "learning_rate": 3.445945945945946e-06,
+      "loss": 0.2082,
+      "step": 153
+    },
+    {
+      "epoch": 0.03469738361449855,
+      "grad_norm": 0.4656449371886705,
+      "learning_rate": 3.468468468468469e-06,
+      "loss": 0.2184,
+      "step": 154
+    },
+    {
+      "epoch": 0.03492269130030698,
+      "grad_norm": 0.5026818550664739,
+      "learning_rate": 3.490990990990991e-06,
+      "loss": 0.2231,
+      "step": 155
+    },
+    {
+      "epoch": 0.03514799898611541,
+      "grad_norm": 0.5322757840474942,
+      "learning_rate": 3.513513513513514e-06,
+      "loss": 0.2458,
+      "step": 156
+    },
+    {
+      "epoch": 0.03537330667192384,
+      "grad_norm": 0.5107332667843115,
+      "learning_rate": 3.536036036036036e-06,
+      "loss": 0.2147,
+      "step": 157
+    },
+    {
+      "epoch": 0.035598614357732275,
+      "grad_norm": 0.5053701639493772,
+      "learning_rate": 3.5585585585585584e-06,
+      "loss": 0.2084,
+      "step": 158
+    },
+    {
+      "epoch": 0.03582392204354071,
+      "grad_norm": 0.48856795307644946,
+      "learning_rate": 3.5810810810810816e-06,
+      "loss": 0.2087,
+      "step": 159
+    },
+    {
+      "epoch": 0.036049229729349146,
+      "grad_norm": 0.5910012852386906,
+      "learning_rate": 3.603603603603604e-06,
+      "loss": 0.2578,
+      "step": 160
+    },
+    {
+      "epoch": 0.03627453741515758,
+      "grad_norm": 0.5206932068291034,
+      "learning_rate": 3.6261261261261266e-06,
+      "loss": 0.228,
+      "step": 161
+    },
+    {
+      "epoch": 0.03649984510096601,
+      "grad_norm": 0.506397979121453,
+      "learning_rate": 3.648648648648649e-06,
+      "loss": 0.2218,
+      "step": 162
+    },
+    {
+      "epoch": 0.03672515278677444,
+      "grad_norm": 0.51976705428227,
+      "learning_rate": 3.6711711711711716e-06,
+      "loss": 0.2035,
+      "step": 163
+    },
+    {
+      "epoch": 0.03695046047258287,
+      "grad_norm": 0.5024417944839559,
+      "learning_rate": 3.693693693693694e-06,
+      "loss": 0.2226,
+      "step": 164
+    },
+    {
+      "epoch": 0.037175768158391305,
+      "grad_norm": 0.5121519754408757,
+      "learning_rate": 3.7162162162162162e-06,
+      "loss": 0.2157,
+      "step": 165
+    },
+    {
+      "epoch": 0.03740107584419974,
+      "grad_norm": 0.48543181216606685,
+      "learning_rate": 3.7387387387387394e-06,
+      "loss": 0.217,
+      "step": 166
+    },
+    {
+      "epoch": 0.03762638353000817,
+      "grad_norm": 0.5253206660875263,
+      "learning_rate": 3.7612612612612612e-06,
+      "loss": 0.2287,
+      "step": 167
+    },
+    {
+      "epoch": 0.0378516912158166,
+      "grad_norm": 0.5519087820882419,
+      "learning_rate": 3.7837837837837844e-06,
+      "loss": 0.2295,
+      "step": 168
+    },
+    {
+      "epoch": 0.03807699890162503,
+      "grad_norm": 0.48023104908811204,
+      "learning_rate": 3.8063063063063067e-06,
+      "loss": 0.2146,
+      "step": 169
+    },
+    {
+      "epoch": 0.038302306587433464,
+      "grad_norm": 0.497406635423104,
+      "learning_rate": 3.828828828828829e-06,
+      "loss": 0.2282,
+      "step": 170
+    },
+    {
+      "epoch": 0.038527614273241896,
+      "grad_norm": 0.5109261032588829,
+      "learning_rate": 3.851351351351352e-06,
+      "loss": 0.2229,
+      "step": 171
+    },
+    {
+      "epoch": 0.03875292195905033,
+      "grad_norm": 0.5105826503113667,
+      "learning_rate": 3.8738738738738744e-06,
+      "loss": 0.2205,
+      "step": 172
+    },
+    {
+      "epoch": 0.03897822964485876,
+      "grad_norm": 0.5062865925051617,
+      "learning_rate": 3.896396396396397e-06,
+      "loss": 0.2106,
+      "step": 173
+    },
+    {
+      "epoch": 0.03920353733066719,
+      "grad_norm": 0.5421073045426972,
+      "learning_rate": 3.918918918918919e-06,
+      "loss": 0.2298,
+      "step": 174
+    },
+    {
+      "epoch": 0.039428845016475624,
+      "grad_norm": 0.5143452971037656,
+      "learning_rate": 3.941441441441442e-06,
+      "loss": 0.2195,
+      "step": 175
+    },
+    {
+      "epoch": 0.039654152702284055,
+      "grad_norm": 0.4951933237019935,
+      "learning_rate": 3.9639639639639645e-06,
+      "loss": 0.2223,
+      "step": 176
+    },
+    {
+      "epoch": 0.03987946038809249,
+      "grad_norm": 0.47561889500434723,
+      "learning_rate": 3.986486486486487e-06,
+      "loss": 0.2021,
+      "step": 177
+    },
+    {
+      "epoch": 0.04010476807390092,
+      "grad_norm": 0.47940239501680004,
+      "learning_rate": 4.009009009009009e-06,
+      "loss": 0.2271,
+      "step": 178
+    },
+    {
+      "epoch": 0.04033007575970935,
+      "grad_norm": 0.5296219636082108,
+      "learning_rate": 4.031531531531531e-06,
+      "loss": 0.2103,
+      "step": 179
+    },
+    {
+      "epoch": 0.04055538344551778,
+      "grad_norm": 0.5344357933055536,
+      "learning_rate": 4.0540540540540545e-06,
+      "loss": 0.2321,
+      "step": 180
+    },
+    {
+      "epoch": 0.040780691131326215,
+      "grad_norm": 0.5454856188785351,
+      "learning_rate": 4.076576576576577e-06,
+      "loss": 0.2377,
+      "step": 181
+    },
+    {
+      "epoch": 0.041005998817134647,
+      "grad_norm": 0.48278341354458393,
+      "learning_rate": 4.099099099099099e-06,
+      "loss": 0.2023,
+      "step": 182
+    },
+    {
+      "epoch": 0.04123130650294308,
+      "grad_norm": 0.4990000028566509,
+      "learning_rate": 4.121621621621622e-06,
+      "loss": 0.1954,
+      "step": 183
+    },
+    {
+      "epoch": 0.04145661418875152,
+      "grad_norm": 0.49432596571018167,
+      "learning_rate": 4.1441441441441446e-06,
+      "loss": 0.2026,
+      "step": 184
+    },
+    {
+      "epoch": 0.04168192187455995,
+      "grad_norm": 0.5323740667881898,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 0.2357,
+      "step": 185
+    },
+    {
+      "epoch": 0.04190722956036838,
+      "grad_norm": 0.49808163776567965,
+      "learning_rate": 4.189189189189189e-06,
+      "loss": 0.2131,
+      "step": 186
+    },
+    {
+      "epoch": 0.04213253724617681,
+      "grad_norm": 0.5274371891475365,
+      "learning_rate": 4.2117117117117115e-06,
+      "loss": 0.2258,
+      "step": 187
+    },
+    {
+      "epoch": 0.042357844931985245,
+      "grad_norm": 0.49290274768513864,
+      "learning_rate": 4.234234234234235e-06,
+      "loss": 0.1998,
+      "step": 188
+    },
+    {
+      "epoch": 0.042583152617793676,
+      "grad_norm": 0.49281647121251326,
+      "learning_rate": 4.256756756756757e-06,
+      "loss": 0.1896,
+      "step": 189
+    },
+    {
+      "epoch": 0.04280846030360211,
+      "grad_norm": 0.45544492366938527,
+      "learning_rate": 4.27927927927928e-06,
+      "loss": 0.2038,
+      "step": 190
+    },
+    {
+      "epoch": 0.04303376798941054,
+      "grad_norm": 0.5004631867414786,
+      "learning_rate": 4.301801801801802e-06,
+      "loss": 0.2127,
+      "step": 191
+    },
+    {
+      "epoch": 0.04325907567521897,
+      "grad_norm": 0.5303953782202576,
+      "learning_rate": 4.324324324324325e-06,
+      "loss": 0.2192,
+      "step": 192
+    },
+    {
+      "epoch": 0.043484383361027404,
+      "grad_norm": 0.4990270446873571,
+      "learning_rate": 4.346846846846847e-06,
+      "loss": 0.2129,
+      "step": 193
+    },
+    {
+      "epoch": 0.043709691046835836,
+      "grad_norm": 0.519415561287903,
+      "learning_rate": 4.369369369369369e-06,
+      "loss": 0.2256,
+      "step": 194
+    },
+    {
+      "epoch": 0.04393499873264427,
+      "grad_norm": 0.534723441521201,
+      "learning_rate": 4.391891891891892e-06,
+      "loss": 0.2062,
+      "step": 195
+    },
+    {
+      "epoch": 0.0441603064184527,
+      "grad_norm": 0.48622823371636226,
+      "learning_rate": 4.414414414414415e-06,
+      "loss": 0.213,
+      "step": 196
+    },
+    {
+      "epoch": 0.04438561410426113,
+      "grad_norm": 0.5119797847730078,
+      "learning_rate": 4.436936936936938e-06,
+      "loss": 0.2139,
+      "step": 197
+    },
+    {
+      "epoch": 0.04461092179006956,
+      "grad_norm": 0.5153856589899163,
+      "learning_rate": 4.45945945945946e-06,
+      "loss": 0.2086,
+      "step": 198
+    },
+    {
+      "epoch": 0.044836229475877995,
+      "grad_norm": 0.5225611573890641,
+      "learning_rate": 4.4819819819819824e-06,
+      "loss": 0.2262,
+      "step": 199
+    },
+    {
+      "epoch": 0.04506153716168643,
+      "grad_norm": 0.5064606336826201,
+      "learning_rate": 4.504504504504505e-06,
+      "loss": 0.2121,
+      "step": 200
+    },
+    {
+      "epoch": 0.04528684484749486,
+      "grad_norm": 0.5200580083558093,
+      "learning_rate": 4.527027027027027e-06,
+      "loss": 0.2174,
+      "step": 201
+    },
+    {
+      "epoch": 0.04551215253330329,
+      "grad_norm": 0.5072035357190301,
+      "learning_rate": 4.54954954954955e-06,
+      "loss": 0.2069,
+      "step": 202
+    },
+    {
+      "epoch": 0.04573746021911172,
+      "grad_norm": 0.5428413804423315,
+      "learning_rate": 4.5720720720720725e-06,
+      "loss": 0.2215,
+      "step": 203
+    },
+    {
+      "epoch": 0.045962767904920154,
+      "grad_norm": 0.5062085074677846,
+      "learning_rate": 4.594594594594596e-06,
+      "loss": 0.2159,
+      "step": 204
+    },
+    {
+      "epoch": 0.046188075590728586,
+      "grad_norm": 0.5077756214561331,
+      "learning_rate": 4.617117117117118e-06,
+      "loss": 0.2255,
+      "step": 205
+    },
+    {
+      "epoch": 0.04641338327653702,
+      "grad_norm": 0.516342590401075,
+      "learning_rate": 4.63963963963964e-06,
+      "loss": 0.2334,
+      "step": 206
+    },
+    {
+      "epoch": 0.04663869096234545,
+      "grad_norm": 0.5146447530347443,
+      "learning_rate": 4.6621621621621625e-06,
+      "loss": 0.2202,
+      "step": 207
+    },
+    {
+      "epoch": 0.04686399864815388,
+      "grad_norm": 0.5218717935866587,
+      "learning_rate": 4.684684684684685e-06,
+      "loss": 0.2227,
+      "step": 208
+    },
+    {
+      "epoch": 0.04708930633396232,
+      "grad_norm": 0.5108508636146013,
+      "learning_rate": 4.707207207207208e-06,
+      "loss": 0.2159,
+      "step": 209
+    },
+    {
+      "epoch": 0.04731461401977075,
+      "grad_norm": 0.49732513414567286,
+      "learning_rate": 4.72972972972973e-06,
+      "loss": 0.2187,
+      "step": 210
+    },
+    {
+      "epoch": 0.047539921705579184,
+      "grad_norm": 0.48464337118429296,
+      "learning_rate": 4.7522522522522526e-06,
+      "loss": 0.1996,
+      "step": 211
+    },
+    {
+      "epoch": 0.047765229391387616,
+      "grad_norm": 0.5456833150991911,
+      "learning_rate": 4.774774774774775e-06,
+      "loss": 0.2249,
+      "step": 212
+    },
+    {
+      "epoch": 0.04799053707719605,
+      "grad_norm": 0.492362483621816,
+      "learning_rate": 4.797297297297297e-06,
+      "loss": 0.1968,
+      "step": 213
+    },
+    {
+      "epoch": 0.04821584476300448,
+      "grad_norm": 0.5257529230994231,
+      "learning_rate": 4.81981981981982e-06,
+      "loss": 0.2097,
+      "step": 214
+    },
+    {
+      "epoch": 0.04844115244881291,
+      "grad_norm": 0.5358862789377798,
+      "learning_rate": 4.842342342342343e-06,
+      "loss": 0.2101,
+      "step": 215
+    },
+    {
+      "epoch": 0.048666460134621344,
+      "grad_norm": 0.5077109074089186,
+      "learning_rate": 4.864864864864866e-06,
+      "loss": 0.2142,
+      "step": 216
+    },
+    {
+      "epoch": 0.048891767820429775,
+      "grad_norm": 0.5048470686894994,
+      "learning_rate": 4.887387387387388e-06,
+      "loss": 0.2033,
+      "step": 217
+    },
+    {
+      "epoch": 0.04911707550623821,
+      "grad_norm": 0.5601972610125039,
+      "learning_rate": 4.90990990990991e-06,
+      "loss": 0.2183,
+      "step": 218
+    },
+    {
+      "epoch": 0.04934238319204664,
+      "grad_norm": 0.5449553809805391,
+      "learning_rate": 4.932432432432433e-06,
+      "loss": 0.2153,
+      "step": 219
+    },
+    {
+      "epoch": 0.04956769087785507,
+      "grad_norm": 0.5751833317104584,
+      "learning_rate": 4.954954954954955e-06,
+      "loss": 0.2315,
+      "step": 220
+    },
+    {
+      "epoch": 0.0497929985636635,
+      "grad_norm": 0.5707106430637009,
+      "learning_rate": 4.977477477477478e-06,
+      "loss": 0.2184,
+      "step": 221
+    },
+    {
+      "epoch": 0.050018306249471935,
+      "grad_norm": 0.5132652534577439,
+      "learning_rate": 5e-06,
+      "loss": 0.216,
+      "step": 222
+    },
+    {
+      "epoch": 0.050243613935280367,
+      "grad_norm": 0.539045624424229,
+      "learning_rate": 4.999999305921593e-06,
+      "loss": 0.2141,
+      "step": 223
+    },
+    {
+      "epoch": 0.0504689216210888,
+      "grad_norm": 0.5075400773617107,
+      "learning_rate": 4.999997223686756e-06,
+      "loss": 0.2013,
+      "step": 224
+    },
+    {
+      "epoch": 0.05069422930689723,
+      "grad_norm": 0.5494829828837551,
+      "learning_rate": 4.9999937532966454e-06,
+      "loss": 0.2347,
+      "step": 225
+    },
+    {
+      "epoch": 0.05091953699270566,
+      "grad_norm": 0.46487989660605866,
+      "learning_rate": 4.999988894753189e-06,
+      "loss": 0.1956,
+      "step": 226
+    },
+    {
+      "epoch": 0.051144844678514094,
+      "grad_norm": 0.5753347536093394,
+      "learning_rate": 4.999982648059082e-06,
+      "loss": 0.2384,
+      "step": 227
+    },
+    {
+      "epoch": 0.051370152364322526,
+      "grad_norm": 0.5651567672972739,
+      "learning_rate": 4.999975013217796e-06,
+      "loss": 0.2252,
+      "step": 228
+    },
+    {
+      "epoch": 0.05159546005013096,
+      "grad_norm": 0.5315878474657139,
+      "learning_rate": 4.99996599023357e-06,
+      "loss": 0.227,
+      "step": 229
+    },
+    {
+      "epoch": 0.05182076773593939,
+      "grad_norm": 0.5177944206907267,
+      "learning_rate": 4.999955579111413e-06,
+      "loss": 0.2147,
+      "step": 230
+    },
+    {
+      "epoch": 0.05204607542174782,
+      "grad_norm": 0.49178836271858745,
+      "learning_rate": 4.999943779857106e-06,
+      "loss": 0.1958,
+      "step": 231
+    },
+    {
+      "epoch": 0.05227138310755625,
+      "grad_norm": 0.5403786614660475,
+      "learning_rate": 4.9999305924772e-06,
+      "loss": 0.2243,
+      "step": 232
+    },
+    {
+      "epoch": 0.05249669079336469,
+      "grad_norm": 0.5249176799168239,
+      "learning_rate": 4.999916016979019e-06,
+      "loss": 0.2067,
+      "step": 233
+    },
+    {
+      "epoch": 0.052721998479173124,
+      "grad_norm": 0.5514864920080108,
+      "learning_rate": 4.999900053370657e-06,
+      "loss": 0.2287,
+      "step": 234
+    },
+    {
+      "epoch": 0.052947306164981556,
+      "grad_norm": 0.5183980485361032,
+      "learning_rate": 4.999882701660975e-06,
+      "loss": 0.2154,
+      "step": 235
+    },
+    {
+      "epoch": 0.05317261385078999,
+      "grad_norm": 0.48926124135397414,
+      "learning_rate": 4.99986396185961e-06,
+      "loss": 0.1996,
+      "step": 236
+    },
+    {
+      "epoch": 0.05339792153659842,
+      "grad_norm": 0.5520877940291486,
+      "learning_rate": 4.999843833976967e-06,
+      "loss": 0.2238,
+      "step": 237
+    },
+    {
+      "epoch": 0.05362322922240685,
+      "grad_norm": 0.48531468109936043,
+      "learning_rate": 4.999822318024222e-06,
+      "loss": 0.197,
+      "step": 238
+    },
+    {
+      "epoch": 0.05384853690821528,
+      "grad_norm": 0.5315551790916936,
+      "learning_rate": 4.999799414013322e-06,
+      "loss": 0.2285,
+      "step": 239
+    },
+    {
+      "epoch": 0.054073844594023715,
+      "grad_norm": 0.4824546627085691,
+      "learning_rate": 4.9997751219569844e-06,
+      "loss": 0.1994,
+      "step": 240
+    },
+    {
+      "epoch": 0.05429915227983215,
+      "grad_norm": 0.48508802248921484,
+      "learning_rate": 4.999749441868699e-06,
+      "loss": 0.1941,
+      "step": 241
+    },
+    {
+      "epoch": 0.05452445996564058,
+      "grad_norm": 0.48336029019618354,
+      "learning_rate": 4.999722373762725e-06,
+      "loss": 0.1964,
+      "step": 242
+    },
+    {
+      "epoch": 0.05474976765144901,
+      "grad_norm": 0.5233609224975063,
+      "learning_rate": 4.9996939176540895e-06,
+      "loss": 0.2101,
+      "step": 243
+    },
+    {
+      "epoch": 0.05497507533725744,
+      "grad_norm": 0.5366150291161705,
+      "learning_rate": 4.999664073558596e-06,
+      "loss": 0.2217,
+      "step": 244
+    },
+    {
+      "epoch": 0.055200383023065874,
+      "grad_norm": 0.5273997275923541,
+      "learning_rate": 4.999632841492815e-06,
+      "loss": 0.2192,
+      "step": 245
+    },
+    {
+      "epoch": 0.055425690708874306,
+      "grad_norm": 0.47872243223030636,
+      "learning_rate": 4.999600221474089e-06,
+      "loss": 0.1949,
+      "step": 246
+    },
+    {
+      "epoch": 0.05565099839468274,
+      "grad_norm": 0.462778258865609,
+      "learning_rate": 4.999566213520529e-06,
+      "loss": 0.1891,
+      "step": 247
+    },
+    {
+      "epoch": 0.05587630608049117,
+      "grad_norm": 0.5192808679253843,
+      "learning_rate": 4.99953081765102e-06,
+      "loss": 0.2025,
+      "step": 248
+    },
+    {
+      "epoch": 0.0561016137662996,
+      "grad_norm": 0.4912780617421906,
+      "learning_rate": 4.999494033885215e-06,
+      "loss": 0.1942,
+      "step": 249
+    },
+    {
+      "epoch": 0.056326921452108034,
+      "grad_norm": 0.5893997616218107,
+      "learning_rate": 4.999455862243539e-06,
+      "loss": 0.2222,
+      "step": 250
+    },
+    {
+      "epoch": 0.056552229137916465,
+      "grad_norm": 0.5167969480025867,
+      "learning_rate": 4.999416302747189e-06,
+      "loss": 0.2159,
+      "step": 251
+    },
+    {
+      "epoch": 0.0567775368237249,
+      "grad_norm": 0.5182504713211056,
+      "learning_rate": 4.999375355418128e-06,
+      "loss": 0.2231,
+      "step": 252
+    },
+    {
+      "epoch": 0.05700284450953333,
+      "grad_norm": 0.5730553068204561,
+      "learning_rate": 4.999333020279094e-06,
+      "loss": 0.2011,
+      "step": 253
+    },
+    {
+      "epoch": 0.05722815219534176,
+      "grad_norm": 0.5219964473354403,
+      "learning_rate": 4.999289297353593e-06,
+      "loss": 0.2123,
+      "step": 254
+    },
+    {
+      "epoch": 0.05745345988115019,
+      "grad_norm": 0.5140790599827941,
+      "learning_rate": 4.9992441866659054e-06,
+      "loss": 0.2115,
+      "step": 255
+    },
+    {
+      "epoch": 0.057678767566958625,
+      "grad_norm": 0.5218855782116167,
+      "learning_rate": 4.999197688241076e-06,
+      "loss": 0.1882,
+      "step": 256
+    },
+    {
+      "epoch": 0.05790407525276706,
+      "grad_norm": 0.4957230593038792,
+      "learning_rate": 4.999149802104926e-06,
+      "loss": 0.1952,
+      "step": 257
+    },
+    {
+      "epoch": 0.058129382938575495,
+      "grad_norm": 0.4874351482139566,
+      "learning_rate": 4.999100528284045e-06,
+      "loss": 0.2074,
+      "step": 258
+    },
+    {
+      "epoch": 0.05835469062438393,
+      "grad_norm": 0.4999889869456212,
+      "learning_rate": 4.999049866805793e-06,
+      "loss": 0.2155,
+      "step": 259
+    },
+    {
+      "epoch": 0.05857999831019236,
+      "grad_norm": 0.495198415318367,
+      "learning_rate": 4.998997817698298e-06,
+      "loss": 0.1936,
+      "step": 260
+    },
+    {
+      "epoch": 0.05880530599600079,
+      "grad_norm": 0.5540297345897838,
+      "learning_rate": 4.998944380990462e-06,
+      "loss": 0.2295,
+      "step": 261
+    },
+    {
+      "epoch": 0.05903061368180922,
+      "grad_norm": 0.5175758124205073,
+      "learning_rate": 4.998889556711958e-06,
+      "loss": 0.2163,
+      "step": 262
+    },
+    {
+      "epoch": 0.059255921367617655,
+      "grad_norm": 0.5276038982841817,
+      "learning_rate": 4.998833344893226e-06,
+      "loss": 0.1934,
+      "step": 263
+    },
+    {
+      "epoch": 0.059481229053426087,
+      "grad_norm": 0.5411162569824867,
+      "learning_rate": 4.998775745565479e-06,
+      "loss": 0.2199,
+      "step": 264
+    },
+    {
+      "epoch": 0.05970653673923452,
+      "grad_norm": 0.5373294969263402,
+      "learning_rate": 4.998716758760701e-06,
+      "loss": 0.2254,
+      "step": 265
+    },
+    {
+      "epoch": 0.05993184442504295,
+      "grad_norm": 0.5141799318273902,
+      "learning_rate": 4.998656384511643e-06,
+      "loss": 0.2102,
+      "step": 266
+    },
+    {
+      "epoch": 0.06015715211085138,
+      "grad_norm": 0.475281124330064,
+      "learning_rate": 4.998594622851829e-06,
+      "loss": 0.1692,
+      "step": 267
+    },
+    {
+      "epoch": 0.060382459796659814,
+      "grad_norm": 0.5380386686045225,
+      "learning_rate": 4.9985314738155545e-06,
+      "loss": 0.2162,
+      "step": 268
+    },
+    {
+      "epoch": 0.060607767482468246,
+      "grad_norm": 0.4897137619103139,
+      "learning_rate": 4.9984669374378825e-06,
+      "loss": 0.1958,
+      "step": 269
+    },
+    {
+      "epoch": 0.06083307516827668,
+      "grad_norm": 0.534719780931734,
+      "learning_rate": 4.9984010137546475e-06,
+      "loss": 0.236,
+      "step": 270
+    },
+    {
+      "epoch": 0.06105838285408511,
+      "grad_norm": 0.5043803991722573,
+      "learning_rate": 4.998333702802457e-06,
+      "loss": 0.2067,
+      "step": 271
+    },
+    {
+      "epoch": 0.06128369053989354,
+      "grad_norm": 0.5300708413018712,
+      "learning_rate": 4.998265004618682e-06,
+      "loss": 0.2178,
+      "step": 272
+    },
+    {
+      "epoch": 0.06150899822570197,
+      "grad_norm": 0.49714976101530717,
+      "learning_rate": 4.998194919241471e-06,
+      "loss": 0.2019,
+      "step": 273
+    },
+    {
+      "epoch": 0.061734305911510405,
+      "grad_norm": 0.5282286817738815,
+      "learning_rate": 4.998123446709739e-06,
+      "loss": 0.2135,
+      "step": 274
+    },
+    {
+      "epoch": 0.06195961359731884,
+      "grad_norm": 0.48789888952975713,
+      "learning_rate": 4.998050587063173e-06,
+      "loss": 0.198,
+      "step": 275
+    },
+    {
+      "epoch": 0.06218492128312727,
+      "grad_norm": 0.44970788932918077,
+      "learning_rate": 4.997976340342226e-06,
+      "loss": 0.2017,
+      "step": 276
+    },
+    {
+      "epoch": 0.0624102289689357,
+      "grad_norm": 0.5583928627478204,
+      "learning_rate": 4.997900706588129e-06,
+      "loss": 0.2158,
+      "step": 277
+    },
+    {
+      "epoch": 0.06263553665474414,
+      "grad_norm": 0.5258990624055785,
+      "learning_rate": 4.997823685842875e-06,
+      "loss": 0.2149,
+      "step": 278
+    },
+    {
+      "epoch": 0.06286084434055257,
+      "grad_norm": 0.46981935210103615,
+      "learning_rate": 4.997745278149233e-06,
+      "loss": 0.1874,
+      "step": 279
+    },
+    {
+      "epoch": 0.063086152026361,
+      "grad_norm": 0.5041562581871978,
+      "learning_rate": 4.997665483550739e-06,
+      "loss": 0.1942,
+      "step": 280
+    },
+    {
+      "epoch": 0.06331145971216944,
+      "grad_norm": 0.5324552273484924,
+      "learning_rate": 4.997584302091699e-06,
+      "loss": 0.2202,
+      "step": 281
+    },
+    {
+      "epoch": 0.06353676739797787,
+      "grad_norm": 0.5478873025934429,
+      "learning_rate": 4.997501733817191e-06,
+      "loss": 0.21,
+      "step": 282
+    },
+    {
+      "epoch": 0.0637620750837863,
+      "grad_norm": 0.46598209735292806,
+      "learning_rate": 4.997417778773064e-06,
+      "loss": 0.1811,
+      "step": 283
+    },
+    {
+      "epoch": 0.06398738276959473,
+      "grad_norm": 0.5996968824573213,
+      "learning_rate": 4.997332437005932e-06,
+      "loss": 0.2384,
+      "step": 284
+    },
+    {
+      "epoch": 0.06421269045540316,
+      "grad_norm": 0.5051133961780844,
+      "learning_rate": 4.9972457085631825e-06,
+      "loss": 0.2043,
+      "step": 285
+    },
+    {
+      "epoch": 0.0644379981412116,
+      "grad_norm": 0.5040193073935666,
+      "learning_rate": 4.997157593492974e-06,
+      "loss": 0.1825,
+      "step": 286
+    },
+    {
+      "epoch": 0.06466330582702003,
+      "grad_norm": 0.4888013540021299,
+      "learning_rate": 4.997068091844233e-06,
+      "loss": 0.2068,
+      "step": 287
+    },
+    {
+      "epoch": 0.06488861351282846,
+      "grad_norm": 0.49915136047539255,
+      "learning_rate": 4.996977203666657e-06,
+      "loss": 0.1984,
+      "step": 288
+    },
+    {
+      "epoch": 0.06511392119863689,
+      "grad_norm": 0.5078177992971098,
+      "learning_rate": 4.99688492901071e-06,
+      "loss": 0.1993,
+      "step": 289
+    },
+    {
+      "epoch": 0.06533922888444532,
+      "grad_norm": 0.5236192779491511,
+      "learning_rate": 4.996791267927632e-06,
+      "loss": 0.2027,
+      "step": 290
+    },
+    {
+      "epoch": 0.06556453657025375,
+      "grad_norm": 0.5245389871527842,
+      "learning_rate": 4.996696220469429e-06,
+      "loss": 0.1979,
+      "step": 291
+    },
+    {
+      "epoch": 0.06578984425606219,
+      "grad_norm": 0.5395585265194126,
+      "learning_rate": 4.996599786688876e-06,
+      "loss": 0.212,
+      "step": 292
+    },
+    {
+      "epoch": 0.06601515194187062,
+      "grad_norm": 0.5353267861924169,
+      "learning_rate": 4.996501966639519e-06,
+      "loss": 0.214,
+      "step": 293
+    },
+    {
+      "epoch": 0.06624045962767905,
+      "grad_norm": 0.5363756637398909,
+      "learning_rate": 4.996402760375676e-06,
+      "loss": 0.2353,
+      "step": 294
+    },
+    {
+      "epoch": 0.06646576731348748,
+      "grad_norm": 0.574596026064565,
+      "learning_rate": 4.99630216795243e-06,
+      "loss": 0.2202,
+      "step": 295
+    },
+    {
+      "epoch": 0.06669107499929591,
+      "grad_norm": 0.5410467145659404,
+      "learning_rate": 4.996200189425638e-06,
+      "loss": 0.2077,
+      "step": 296
+    },
+    {
+      "epoch": 0.06691638268510434,
+      "grad_norm": 0.4998899935289798,
+      "learning_rate": 4.996096824851923e-06,
+      "loss": 0.2141,
+      "step": 297
+    },
+    {
+      "epoch": 0.06714169037091278,
+      "grad_norm": 0.5162735905138214,
+      "learning_rate": 4.9959920742886815e-06,
+      "loss": 0.2006,
+      "step": 298
+    },
+    {
+      "epoch": 0.06736699805672121,
+      "grad_norm": 0.5737456586581821,
+      "learning_rate": 4.9958859377940765e-06,
+      "loss": 0.2027,
+      "step": 299
+    },
+    {
+      "epoch": 0.06759230574252964,
+      "grad_norm": 0.5413621557656281,
+      "learning_rate": 4.995778415427042e-06,
+      "loss": 0.2098,
+      "step": 300
+    },
+    {
+      "epoch": 0.06781761342833807,
+      "grad_norm": 0.577394779919834,
+      "learning_rate": 4.99566950724728e-06,
+      "loss": 0.2269,
+      "step": 301
+    },
+    {
+      "epoch": 0.0680429211141465,
+      "grad_norm": 0.515822079782068,
+      "learning_rate": 4.995559213315267e-06,
+      "loss": 0.2261,
+      "step": 302
+    },
+    {
+      "epoch": 0.06826822879995494,
+      "grad_norm": 0.4712268338728312,
+      "learning_rate": 4.995447533692239e-06,
+      "loss": 0.2003,
+      "step": 303
+    },
+    {
+      "epoch": 0.06849353648576337,
+      "grad_norm": 0.5134784305459534,
+      "learning_rate": 4.995334468440213e-06,
+      "loss": 0.1997,
+      "step": 304
+    },
+    {
+      "epoch": 0.0687188441715718,
+      "grad_norm": 0.4907228514504316,
+      "learning_rate": 4.995220017621967e-06,
+      "loss": 0.1896,
+      "step": 305
+    },
+    {
+      "epoch": 0.06894415185738023,
+      "grad_norm": 0.4962154271689104,
+      "learning_rate": 4.995104181301052e-06,
+      "loss": 0.2027,
+      "step": 306
+    },
+    {
+      "epoch": 0.06916945954318866,
+      "grad_norm": 0.5537049140921233,
+      "learning_rate": 4.994986959541788e-06,
+      "loss": 0.213,
+      "step": 307
+    },
+    {
+      "epoch": 0.0693947672289971,
+      "grad_norm": 0.49941255076713437,
+      "learning_rate": 4.994868352409263e-06,
+      "loss": 0.1964,
+      "step": 308
+    },
+    {
+      "epoch": 0.06962007491480553,
+      "grad_norm": 0.5054354369750879,
+      "learning_rate": 4.994748359969336e-06,
+      "loss": 0.2071,
+      "step": 309
+    },
+    {
+      "epoch": 0.06984538260061396,
+      "grad_norm": 0.5172274273450349,
+      "learning_rate": 4.9946269822886335e-06,
+      "loss": 0.2034,
+      "step": 310
+    },
+    {
+      "epoch": 0.07007069028642239,
+      "grad_norm": 0.4931991542019778,
+      "learning_rate": 4.994504219434553e-06,
+      "loss": 0.2136,
+      "step": 311
+    },
+    {
+      "epoch": 0.07029599797223082,
+      "grad_norm": 0.5264002662303077,
+      "learning_rate": 4.9943800714752586e-06,
+      "loss": 0.2181,
+      "step": 312
+    },
+    {
+      "epoch": 0.07052130565803925,
+      "grad_norm": 0.5062978313923276,
+      "learning_rate": 4.994254538479687e-06,
+      "loss": 0.2001,
+      "step": 313
+    },
+    {
+      "epoch": 0.07074661334384769,
+      "grad_norm": 0.5508272404060727,
+      "learning_rate": 4.9941276205175405e-06,
+      "loss": 0.209,
+      "step": 314
+    },
+    {
+      "epoch": 0.07097192102965612,
+      "grad_norm": 0.5076274602102792,
+      "learning_rate": 4.993999317659293e-06,
+      "loss": 0.2038,
+      "step": 315
+    },
+    {
+      "epoch": 0.07119722871546455,
+      "grad_norm": 0.5122198034692836,
+      "learning_rate": 4.9938696299761856e-06,
+      "loss": 0.1889,
+      "step": 316
+    },
+    {
+      "epoch": 0.07142253640127298,
+      "grad_norm": 0.6066990312387696,
+      "learning_rate": 4.9937385575402284e-06,
+      "loss": 0.2387,
+      "step": 317
+    },
+    {
+      "epoch": 0.07164784408708141,
+      "grad_norm": 0.5113729441842336,
+      "learning_rate": 4.993606100424202e-06,
+      "loss": 0.1925,
+      "step": 318
+    },
+    {
+      "epoch": 0.07187315177288986,
+      "grad_norm": 0.5434767861340162,
+      "learning_rate": 4.9934722587016555e-06,
+      "loss": 0.2083,
+      "step": 319
+    },
+    {
+      "epoch": 0.07209845945869829,
+      "grad_norm": 0.5220970910896368,
+      "learning_rate": 4.9933370324469045e-06,
+      "loss": 0.2127,
+      "step": 320
+    },
+    {
+      "epoch": 0.07232376714450672,
+      "grad_norm": 0.5133695423087399,
+      "learning_rate": 4.993200421735037e-06,
+      "loss": 0.2072,
+      "step": 321
+    },
+    {
+      "epoch": 0.07254907483031516,
+      "grad_norm": 0.5107258940556315,
+      "learning_rate": 4.993062426641906e-06,
+      "loss": 0.2036,
+      "step": 322
+    },
+    {
+      "epoch": 0.07277438251612359,
+      "grad_norm": 0.5455096340183896,
+      "learning_rate": 4.992923047244136e-06,
+      "loss": 0.2219,
+      "step": 323
+    },
+    {
+      "epoch": 0.07299969020193202,
+      "grad_norm": 0.5218309849291067,
+      "learning_rate": 4.9927822836191185e-06,
+      "loss": 0.2235,
+      "step": 324
+    },
+    {
+      "epoch": 0.07322499788774045,
+      "grad_norm": 0.5057153674679081,
+      "learning_rate": 4.992640135845016e-06,
+      "loss": 0.2078,
+      "step": 325
+    },
+    {
+      "epoch": 0.07345030557354888,
+      "grad_norm": 0.4632931130727815,
+      "learning_rate": 4.992496604000756e-06,
+      "loss": 0.1917,
+      "step": 326
+    },
+    {
+      "epoch": 0.07367561325935731,
+      "grad_norm": 0.5133401143389513,
+      "learning_rate": 4.992351688166038e-06,
+      "loss": 0.1942,
+      "step": 327
+    },
+    {
+      "epoch": 0.07390092094516575,
+      "grad_norm": 0.5425225307579951,
+      "learning_rate": 4.992205388421326e-06,
+      "loss": 0.2136,
+      "step": 328
+    },
+    {
+      "epoch": 0.07412622863097418,
+      "grad_norm": 0.4938836282672529,
+      "learning_rate": 4.992057704847858e-06,
+      "loss": 0.2071,
+      "step": 329
+    },
+    {
+      "epoch": 0.07435153631678261,
+      "grad_norm": 0.5042566721615249,
+      "learning_rate": 4.991908637527634e-06,
+      "loss": 0.2048,
+      "step": 330
+    },
+    {
+      "epoch": 0.07457684400259104,
+      "grad_norm": 0.5079880579268699,
+      "learning_rate": 4.9917581865434275e-06,
+      "loss": 0.192,
+      "step": 331
+    },
+    {
+      "epoch": 0.07480215168839947,
+      "grad_norm": 0.5111920959241932,
+      "learning_rate": 4.9916063519787775e-06,
+      "loss": 0.2007,
+      "step": 332
+    },
+    {
+      "epoch": 0.0750274593742079,
+      "grad_norm": 0.5033017411516871,
+      "learning_rate": 4.991453133917993e-06,
+      "loss": 0.2049,
+      "step": 333
+    },
+    {
+      "epoch": 0.07525276706001634,
+      "grad_norm": 0.5129775312572191,
+      "learning_rate": 4.991298532446149e-06,
+      "loss": 0.2069,
+      "step": 334
+    },
+    {
+      "epoch": 0.07547807474582477,
+      "grad_norm": 0.548652704267977,
+      "learning_rate": 4.991142547649091e-06,
+      "loss": 0.1932,
+      "step": 335
+    },
+    {
+      "epoch": 0.0757033824316332,
+      "grad_norm": 0.4673398399684016,
+      "learning_rate": 4.990985179613431e-06,
+      "loss": 0.1953,
+      "step": 336
+    },
+    {
+      "epoch": 0.07592869011744163,
+      "grad_norm": 0.5295523972404997,
+      "learning_rate": 4.990826428426549e-06,
+      "loss": 0.2049,
+      "step": 337
+    },
+    {
+      "epoch": 0.07615399780325006,
+      "grad_norm": 0.5189619606719836,
+      "learning_rate": 4.990666294176596e-06,
+      "loss": 0.2045,
+      "step": 338
+    },
+    {
+      "epoch": 0.0763793054890585,
+      "grad_norm": 0.5067351401379646,
+      "learning_rate": 4.9905047769524855e-06,
+      "loss": 0.2198,
+      "step": 339
+    },
+    {
+      "epoch": 0.07660461317486693,
+      "grad_norm": 0.5121777968541428,
+      "learning_rate": 4.990341876843904e-06,
+      "loss": 0.1985,
+      "step": 340
+    },
+    {
+      "epoch": 0.07682992086067536,
+      "grad_norm": 0.5311438794688305,
+      "learning_rate": 4.990177593941303e-06,
+      "loss": 0.2059,
+      "step": 341
+    },
+    {
+      "epoch": 0.07705522854648379,
+      "grad_norm": 0.49129677532414406,
+      "learning_rate": 4.9900119283359025e-06,
+      "loss": 0.1895,
+      "step": 342
+    },
+    {
+      "epoch": 0.07728053623229222,
+      "grad_norm": 0.4893539747998295,
+      "learning_rate": 4.989844880119692e-06,
+      "loss": 0.1807,
+      "step": 343
+    },
+    {
+      "epoch": 0.07750584391810066,
+      "grad_norm": 0.5072252752550346,
+      "learning_rate": 4.989676449385426e-06,
+      "loss": 0.2018,
+      "step": 344
+    },
+    {
+      "epoch": 0.07773115160390909,
+      "grad_norm": 0.4940421937088313,
+      "learning_rate": 4.989506636226626e-06,
+      "loss": 0.1967,
+      "step": 345
+    },
+    {
+      "epoch": 0.07795645928971752,
+      "grad_norm": 0.4786128739759888,
+      "learning_rate": 4.989335440737587e-06,
+      "loss": 0.1883,
+      "step": 346
+    },
+    {
+      "epoch": 0.07818176697552595,
+      "grad_norm": 0.4980537115424014,
+      "learning_rate": 4.989162863013364e-06,
+      "loss": 0.1878,
+      "step": 347
+    },
+    {
+      "epoch": 0.07840707466133438,
+      "grad_norm": 0.4874848708154177,
+      "learning_rate": 4.988988903149784e-06,
+      "loss": 0.1909,
+      "step": 348
+    },
+    {
+      "epoch": 0.07863238234714282,
+      "grad_norm": 0.5228954597731107,
+      "learning_rate": 4.9888135612434415e-06,
+      "loss": 0.1974,
+      "step": 349
+    },
+    {
+      "epoch": 0.07885769003295125,
+      "grad_norm": 0.5121660056443094,
+      "learning_rate": 4.988636837391696e-06,
+      "loss": 0.1882,
+      "step": 350
+    },
+    {
+      "epoch": 0.07908299771875968,
+      "grad_norm": 0.5031256954337335,
+      "learning_rate": 4.9884587316926765e-06,
+      "loss": 0.2116,
+      "step": 351
+    },
+    {
+      "epoch": 0.07930830540456811,
+      "grad_norm": 0.5144091853279764,
+      "learning_rate": 4.988279244245278e-06,
+      "loss": 0.2033,
+      "step": 352
+    },
+    {
+      "epoch": 0.07953361309037654,
+      "grad_norm": 0.531214217524162,
+      "learning_rate": 4.988098375149163e-06,
+      "loss": 0.198,
+      "step": 353
+    },
+    {
+      "epoch": 0.07975892077618497,
+      "grad_norm": 0.4889296327628917,
+      "learning_rate": 4.987916124504761e-06,
+      "loss": 0.1997,
+      "step": 354
+    },
+    {
+      "epoch": 0.0799842284619934,
+      "grad_norm": 0.49524804048725096,
+      "learning_rate": 4.987732492413271e-06,
+      "loss": 0.1855,
+      "step": 355
+    },
+    {
+      "epoch": 0.08020953614780184,
+      "grad_norm": 0.4864717158203817,
+      "learning_rate": 4.987547478976655e-06,
+      "loss": 0.1911,
+      "step": 356
+    },
+    {
+      "epoch": 0.08043484383361027,
+      "grad_norm": 0.4936470667020924,
+      "learning_rate": 4.987361084297645e-06,
+      "loss": 0.1968,
+      "step": 357
+    },
+    {
+      "epoch": 0.0806601515194187,
+      "grad_norm": 0.5020980156378522,
+      "learning_rate": 4.987173308479738e-06,
+      "loss": 0.1922,
+      "step": 358
+    },
+    {
+      "epoch": 0.08088545920522713,
+      "grad_norm": 0.467049169764784,
+      "learning_rate": 4.9869841516272004e-06,
+      "loss": 0.1933,
+      "step": 359
+    },
+    {
+      "epoch": 0.08111076689103557,
+      "grad_norm": 0.4926168450953013,
+      "learning_rate": 4.9867936138450635e-06,
+      "loss": 0.1875,
+      "step": 360
+    },
+    {
+      "epoch": 0.081336074576844,
+      "grad_norm": 0.5553008487726967,
+      "learning_rate": 4.986601695239125e-06,
+      "loss": 0.2116,
+      "step": 361
+    },
+    {
+      "epoch": 0.08156138226265243,
+      "grad_norm": 0.46531354659798435,
+      "learning_rate": 4.98640839591595e-06,
+      "loss": 0.2005,
+      "step": 362
+    },
+    {
+      "epoch": 0.08178668994846086,
+      "grad_norm": 0.4975006830434269,
+      "learning_rate": 4.986213715982873e-06,
+      "loss": 0.1997,
+      "step": 363
+    },
+    {
+      "epoch": 0.08201199763426929,
+      "grad_norm": 0.5484910187073875,
+      "learning_rate": 4.986017655547989e-06,
+      "loss": 0.2052,
+      "step": 364
+    },
+    {
+      "epoch": 0.08223730532007772,
+      "grad_norm": 0.4917366895315546,
+      "learning_rate": 4.985820214720165e-06,
+      "loss": 0.1963,
+      "step": 365
+    },
+    {
+      "epoch": 0.08246261300588616,
+      "grad_norm": 0.5079414756556945,
+      "learning_rate": 4.985621393609032e-06,
+      "loss": 0.2171,
+      "step": 366
+    },
+    {
+      "epoch": 0.08268792069169459,
+      "grad_norm": 0.5025287311574784,
+      "learning_rate": 4.98542119232499e-06,
+      "loss": 0.1847,
+      "step": 367
+    },
+    {
+      "epoch": 0.08291322837750303,
+      "grad_norm": 0.5054672772723833,
+      "learning_rate": 4.9852196109792e-06,
+      "loss": 0.1828,
+      "step": 368
+    },
+    {
+      "epoch": 0.08313853606331147,
+      "grad_norm": 0.5119523083386188,
+      "learning_rate": 4.985016649683594e-06,
+      "loss": 0.2057,
+      "step": 369
+    },
+    {
+      "epoch": 0.0833638437491199,
+      "grad_norm": 0.4973589633933691,
+      "learning_rate": 4.984812308550869e-06,
+      "loss": 0.2077,
+      "step": 370
+    },
+    {
+      "epoch": 0.08358915143492833,
+      "grad_norm": 0.4809212159374925,
+      "learning_rate": 4.984606587694488e-06,
+      "loss": 0.1875,
+      "step": 371
+    },
+    {
+      "epoch": 0.08381445912073676,
+      "grad_norm": 0.5694307394026895,
+      "learning_rate": 4.98439948722868e-06,
+      "loss": 0.2076,
+      "step": 372
+    },
+    {
+      "epoch": 0.0840397668065452,
+      "grad_norm": 0.49181225929988026,
+      "learning_rate": 4.9841910072684406e-06,
+      "loss": 0.2081,
+      "step": 373
+    },
+    {
+      "epoch": 0.08426507449235363,
+      "grad_norm": 0.5163094079454916,
+      "learning_rate": 4.98398114792953e-06,
+      "loss": 0.1999,
+      "step": 374
+    },
+    {
+      "epoch": 0.08449038217816206,
+      "grad_norm": 0.48034370172395746,
+      "learning_rate": 4.9837699093284765e-06,
+      "loss": 0.1954,
+      "step": 375
+    },
+    {
+      "epoch": 0.08471568986397049,
+      "grad_norm": 0.5167164355540542,
+      "learning_rate": 4.983557291582572e-06,
+      "loss": 0.2055,
+      "step": 376
+    },
+    {
+      "epoch": 0.08494099754977892,
+      "grad_norm": 0.5221710536395783,
+      "learning_rate": 4.983343294809875e-06,
+      "loss": 0.187,
+      "step": 377
+    },
+    {
+      "epoch": 0.08516630523558735,
+      "grad_norm": 0.5195789727627067,
+      "learning_rate": 4.9831279191292114e-06,
+      "loss": 0.196,
+      "step": 378
+    },
+    {
+      "epoch": 0.08539161292139578,
+      "grad_norm": 0.4749341701034613,
+      "learning_rate": 4.98291116466017e-06,
+      "loss": 0.1979,
+      "step": 379
+    },
+    {
+      "epoch": 0.08561692060720422,
+      "grad_norm": 0.4940468961257005,
+      "learning_rate": 4.982693031523107e-06,
+      "loss": 0.1725,
+      "step": 380
+    },
+    {
+      "epoch": 0.08584222829301265,
+      "grad_norm": 0.49776133813703377,
+      "learning_rate": 4.982473519839144e-06,
+      "loss": 0.1968,
+      "step": 381
+    },
+    {
+      "epoch": 0.08606753597882108,
+      "grad_norm": 0.5406461195372502,
+      "learning_rate": 4.982252629730167e-06,
+      "loss": 0.2002,
+      "step": 382
+    },
+    {
+      "epoch": 0.08629284366462951,
+      "grad_norm": 0.5102425028271348,
+      "learning_rate": 4.982030361318827e-06,
+      "loss": 0.1943,
+      "step": 383
+    },
+    {
+      "epoch": 0.08651815135043794,
+      "grad_norm": 0.4743565750135506,
+      "learning_rate": 4.981806714728543e-06,
+      "loss": 0.1866,
+      "step": 384
+    },
+    {
+      "epoch": 0.08674345903624638,
+      "grad_norm": 0.46885833113908765,
+      "learning_rate": 4.981581690083498e-06,
+      "loss": 0.1726,
+      "step": 385
+    },
+    {
+      "epoch": 0.08696876672205481,
+      "grad_norm": 0.48609707569076405,
+      "learning_rate": 4.981355287508638e-06,
+      "loss": 0.1757,
+      "step": 386
+    },
+    {
+      "epoch": 0.08719407440786324,
+      "grad_norm": 0.4667625682253398,
+      "learning_rate": 4.981127507129677e-06,
+      "loss": 0.1909,
+      "step": 387
+    },
+    {
+      "epoch": 0.08741938209367167,
+      "grad_norm": 0.5218958233886627,
+      "learning_rate": 4.980898349073094e-06,
+      "loss": 0.2007,
+      "step": 388
+    },
+    {
+      "epoch": 0.0876446897794801,
+      "grad_norm": 0.4673222943348308,
+      "learning_rate": 4.9806678134661295e-06,
+      "loss": 0.1915,
+      "step": 389
+    },
+    {
+      "epoch": 0.08786999746528854,
+      "grad_norm": 0.49616638111195194,
+      "learning_rate": 4.980435900436793e-06,
+      "loss": 0.1875,
+      "step": 390
+    },
+    {
+      "epoch": 0.08809530515109697,
+      "grad_norm": 0.4817346343815341,
+      "learning_rate": 4.980202610113857e-06,
+      "loss": 0.1905,
+      "step": 391
+    },
+    {
+      "epoch": 0.0883206128369054,
+      "grad_norm": 0.5351889949883639,
+      "learning_rate": 4.9799679426268575e-06,
+      "loss": 0.2208,
+      "step": 392
+    },
+    {
+      "epoch": 0.08854592052271383,
+      "grad_norm": 0.5047966627283784,
+      "learning_rate": 4.9797318981061e-06,
+      "loss": 0.2039,
+      "step": 393
+    },
+    {
+      "epoch": 0.08877122820852226,
+      "grad_norm": 0.4923828536733769,
+      "learning_rate": 4.979494476682647e-06,
+      "loss": 0.2018,
+      "step": 394
+    },
+    {
+      "epoch": 0.0889965358943307,
+      "grad_norm": 0.48812777672235563,
+      "learning_rate": 4.979255678488332e-06,
+      "loss": 0.1871,
+      "step": 395
+    },
+    {
+      "epoch": 0.08922184358013913,
+      "grad_norm": 0.48533989253285686,
+      "learning_rate": 4.979015503655751e-06,
+      "loss": 0.2001,
+      "step": 396
+    },
+    {
+      "epoch": 0.08944715126594756,
+      "grad_norm": 0.500758368828542,
+      "learning_rate": 4.978773952318263e-06,
+      "loss": 0.2083,
+      "step": 397
+    },
+    {
+      "epoch": 0.08967245895175599,
+      "grad_norm": 0.4751863671531154,
+      "learning_rate": 4.978531024609994e-06,
+      "loss": 0.2025,
+      "step": 398
+    },
+    {
+      "epoch": 0.08989776663756442,
+      "grad_norm": 0.5028504445052441,
+      "learning_rate": 4.978286720665832e-06,
+      "loss": 0.2034,
+      "step": 399
+    },
+    {
+      "epoch": 0.09012307432337285,
+      "grad_norm": 0.4840968995820041,
+      "learning_rate": 4.978041040621428e-06,
+      "loss": 0.1908,
+      "step": 400
+    },
+    {
+      "epoch": 0.09034838200918129,
+      "grad_norm": 0.49160319980568024,
+      "learning_rate": 4.977793984613202e-06,
+      "loss": 0.1944,
+      "step": 401
+    },
+    {
+      "epoch": 0.09057368969498972,
+      "grad_norm": 0.4973629969371634,
+      "learning_rate": 4.977545552778333e-06,
+      "loss": 0.1962,
+      "step": 402
+    },
+    {
+      "epoch": 0.09079899738079815,
+      "grad_norm": 0.47691284590905275,
+      "learning_rate": 4.977295745254766e-06,
+      "loss": 0.1914,
+      "step": 403
+    },
+    {
+      "epoch": 0.09102430506660658,
+      "grad_norm": 0.47003075599798927,
+      "learning_rate": 4.977044562181212e-06,
+      "loss": 0.1847,
+      "step": 404
+    },
+    {
+      "epoch": 0.09124961275241501,
+      "grad_norm": 0.476135794675388,
+      "learning_rate": 4.9767920036971406e-06,
+      "loss": 0.1924,
+      "step": 405
+    },
+    {
+      "epoch": 0.09147492043822344,
+      "grad_norm": 0.4801530579110471,
+      "learning_rate": 4.9765380699427905e-06,
+      "loss": 0.1794,
+      "step": 406
+    },
+    {
+      "epoch": 0.09170022812403188,
+      "grad_norm": 0.5585643748818848,
+      "learning_rate": 4.97628276105916e-06,
+      "loss": 0.2114,
+      "step": 407
+    },
+    {
+      "epoch": 0.09192553580984031,
+      "grad_norm": 0.49406470093699134,
+      "learning_rate": 4.976026077188013e-06,
+      "loss": 0.1964,
+      "step": 408
+    },
+    {
+      "epoch": 0.09215084349564874,
+      "grad_norm": 0.48542519989421573,
+      "learning_rate": 4.975768018471877e-06,
+      "loss": 0.1783,
+      "step": 409
+    },
+    {
+      "epoch": 0.09237615118145717,
+      "grad_norm": 0.48764410824033644,
+      "learning_rate": 4.9755085850540426e-06,
+      "loss": 0.198,
+      "step": 410
+    },
+    {
+      "epoch": 0.0926014588672656,
+      "grad_norm": 0.504358514720316,
+      "learning_rate": 4.9752477770785625e-06,
+      "loss": 0.201,
+      "step": 411
+    },
+    {
+      "epoch": 0.09282676655307404,
+      "grad_norm": 0.49485581002695883,
+      "learning_rate": 4.974985594690255e-06,
+      "loss": 0.1972,
+      "step": 412
+    },
+    {
+      "epoch": 0.09305207423888247,
+      "grad_norm": 0.4770605027062997,
+      "learning_rate": 4.9747220380346975e-06,
+      "loss": 0.1769,
+      "step": 413
+    },
+    {
+      "epoch": 0.0932773819246909,
+      "grad_norm": 0.5054068132119154,
+      "learning_rate": 4.9744571072582365e-06,
+      "loss": 0.2051,
+      "step": 414
+    },
+    {
+      "epoch": 0.09350268961049933,
+      "grad_norm": 0.5064939644802016,
+      "learning_rate": 4.974190802507977e-06,
+      "loss": 0.1935,
+      "step": 415
+    },
+    {
+      "epoch": 0.09372799729630776,
+      "grad_norm": 0.4889748679569221,
+      "learning_rate": 4.973923123931786e-06,
+      "loss": 0.191,
+      "step": 416
+    },
+    {
+      "epoch": 0.09395330498211621,
+      "grad_norm": 0.5089835583101964,
+      "learning_rate": 4.973654071678299e-06,
+      "loss": 0.1927,
+      "step": 417
+    },
+    {
+      "epoch": 0.09417861266792464,
+      "grad_norm": 0.5729086031783864,
+      "learning_rate": 4.973383645896908e-06,
+      "loss": 0.1983,
+      "step": 418
+    },
+    {
+      "epoch": 0.09440392035373307,
+      "grad_norm": 0.4816122719585705,
+      "learning_rate": 4.973111846737772e-06,
+      "loss": 0.1841,
+      "step": 419
+    },
+    {
+      "epoch": 0.0946292280395415,
+      "grad_norm": 0.5084418909549541,
+      "learning_rate": 4.97283867435181e-06,
+      "loss": 0.1868,
+      "step": 420
+    },
+    {
+      "epoch": 0.09485453572534994,
+      "grad_norm": 0.5105786107510725,
+      "learning_rate": 4.972564128890704e-06,
+      "loss": 0.2074,
+      "step": 421
+    },
+    {
+      "epoch": 0.09507984341115837,
+      "grad_norm": 0.48927251560238677,
+      "learning_rate": 4.972288210506902e-06,
+      "loss": 0.1945,
+      "step": 422
+    },
+    {
+      "epoch": 0.0953051510969668,
+      "grad_norm": 0.4939682917412821,
+      "learning_rate": 4.972010919353606e-06,
+      "loss": 0.1901,
+      "step": 423
+    },
+    {
+      "epoch": 0.09553045878277523,
+      "grad_norm": 0.49015495094460076,
+      "learning_rate": 4.971732255584789e-06,
+      "loss": 0.1925,
+      "step": 424
+    },
+    {
+      "epoch": 0.09575576646858366,
+      "grad_norm": 0.4959299200961116,
+      "learning_rate": 4.971452219355182e-06,
+      "loss": 0.2026,
+      "step": 425
+    },
+    {
+      "epoch": 0.0959810741543921,
+      "grad_norm": 0.5064117553090074,
+      "learning_rate": 4.971170810820279e-06,
+      "loss": 0.192,
+      "step": 426
+    },
+    {
+      "epoch": 0.09620638184020053,
+      "grad_norm": 0.49162239980799377,
+      "learning_rate": 4.970888030136335e-06,
+      "loss": 0.1896,
+      "step": 427
+    },
+    {
+      "epoch": 0.09643168952600896,
+      "grad_norm": 0.5188915261797445,
+      "learning_rate": 4.970603877460367e-06,
+      "loss": 0.2043,
+      "step": 428
+    },
+    {
+      "epoch": 0.09665699721181739,
+      "grad_norm": 0.5248312270545016,
+      "learning_rate": 4.970318352950155e-06,
+      "loss": 0.2105,
+      "step": 429
+    },
+    {
+      "epoch": 0.09688230489762582,
+      "grad_norm": 0.4947470788962499,
+      "learning_rate": 4.970031456764242e-06,
+      "loss": 0.2016,
+      "step": 430
+    },
+    {
+      "epoch": 0.09710761258343426,
+      "grad_norm": 0.4725899497424823,
+      "learning_rate": 4.9697431890619265e-06,
+      "loss": 0.1892,
+      "step": 431
+    },
+    {
+      "epoch": 0.09733292026924269,
+      "grad_norm": 0.5106782555738607,
+      "learning_rate": 4.969453550003277e-06,
+      "loss": 0.1916,
+      "step": 432
+    },
+    {
+      "epoch": 0.09755822795505112,
+      "grad_norm": 0.5277102908599764,
+      "learning_rate": 4.969162539749117e-06,
+      "loss": 0.1952,
+      "step": 433
+    },
+    {
+      "epoch": 0.09778353564085955,
+      "grad_norm": 0.47760908967624405,
+      "learning_rate": 4.9688701584610345e-06,
+      "loss": 0.1831,
+      "step": 434
+    },
+    {
+      "epoch": 0.09800884332666798,
+      "grad_norm": 0.44518578891497956,
+      "learning_rate": 4.968576406301377e-06,
+      "loss": 0.1836,
+      "step": 435
+    },
+    {
+      "epoch": 0.09823415101247641,
+      "grad_norm": 0.47596718675919764,
+      "learning_rate": 4.968281283433256e-06,
+      "loss": 0.1927,
+      "step": 436
+    },
+    {
+      "epoch": 0.09845945869828485,
+      "grad_norm": 0.4776725350497191,
+      "learning_rate": 4.96798479002054e-06,
+      "loss": 0.1812,
+      "step": 437
+    },
+    {
+      "epoch": 0.09868476638409328,
+      "grad_norm": 0.5053833290898005,
+      "learning_rate": 4.967686926227862e-06,
+      "loss": 0.1921,
+      "step": 438
+    },
+    {
+      "epoch": 0.09891007406990171,
+      "grad_norm": 0.5190307175787031,
+      "learning_rate": 4.967387692220615e-06,
+      "loss": 0.2063,
+      "step": 439
+    },
+    {
+      "epoch": 0.09913538175571014,
+      "grad_norm": 0.47988599552415206,
+      "learning_rate": 4.967087088164951e-06,
+      "loss": 0.1887,
+      "step": 440
+    },
+    {
+      "epoch": 0.09936068944151857,
+      "grad_norm": 0.4763852121519757,
+      "learning_rate": 4.966785114227785e-06,
+      "loss": 0.1871,
+      "step": 441
+    },
+    {
+      "epoch": 0.099585997127327,
+      "grad_norm": 0.5291626373728735,
+      "learning_rate": 4.966481770576793e-06,
+      "loss": 0.1982,
+      "step": 442
+    },
+    {
+      "epoch": 0.09981130481313544,
+      "grad_norm": 0.4999379174851594,
+      "learning_rate": 4.966177057380409e-06,
+      "loss": 0.186,
+      "step": 443
+    },
+    {
+      "epoch": 0.10003661249894387,
+      "grad_norm": 0.5127885084474579,
+      "learning_rate": 4.965870974807829e-06,
+      "loss": 0.1945,
+      "step": 444
+    },
+    {
+      "epoch": 0.1002619201847523,
+      "grad_norm": 0.5302768350911209,
+      "learning_rate": 4.96556352302901e-06,
+      "loss": 0.2085,
+      "step": 445
+    },
+    {
+      "epoch": 0.10048722787056073,
+      "grad_norm": 0.5240819112362862,
+      "learning_rate": 4.965254702214668e-06,
+      "loss": 0.1919,
+      "step": 446
+    },
+    {
+      "epoch": 0.10071253555636916,
+      "grad_norm": 0.4954501664107016,
+      "learning_rate": 4.96494451253628e-06,
+      "loss": 0.1869,
+      "step": 447
+    },
+    {
+      "epoch": 0.1009378432421776,
+      "grad_norm": 0.5242909937666382,
+      "learning_rate": 4.964632954166081e-06,
+      "loss": 0.2166,
+      "step": 448
+    },
+    {
+      "epoch": 0.10116315092798603,
+      "grad_norm": 0.4544684683438984,
+      "learning_rate": 4.964320027277071e-06,
+      "loss": 0.1664,
+      "step": 449
+    },
+    {
+      "epoch": 0.10138845861379446,
+      "grad_norm": 0.49379164665615777,
+      "learning_rate": 4.964005732043003e-06,
+      "loss": 0.1989,
+      "step": 450
+    },
+    {
+      "epoch": 0.10161376629960289,
+      "grad_norm": 0.4942749092145873,
+      "learning_rate": 4.963690068638397e-06,
+      "loss": 0.191,
+      "step": 451
+    },
+    {
+      "epoch": 0.10183907398541132,
+      "grad_norm": 0.5246799570633436,
+      "learning_rate": 4.963373037238527e-06,
+      "loss": 0.1971,
+      "step": 452
+    },
+    {
+      "epoch": 0.10206438167121976,
+      "grad_norm": 0.5302756608759575,
+      "learning_rate": 4.963054638019429e-06,
+      "loss": 0.1921,
+      "step": 453
+    },
+    {
+      "epoch": 0.10228968935702819,
+      "grad_norm": 0.49314602619260695,
+      "learning_rate": 4.9627348711578996e-06,
+      "loss": 0.1813,
+      "step": 454
+    },
+    {
+      "epoch": 0.10251499704283662,
+      "grad_norm": 0.5039489074404687,
+      "learning_rate": 4.962413736831491e-06,
+      "loss": 0.2017,
+      "step": 455
+    },
+    {
+      "epoch": 0.10274030472864505,
+      "grad_norm": 0.5533083712534903,
+      "learning_rate": 4.962091235218518e-06,
+      "loss": 0.2014,
+      "step": 456
+    },
+    {
+      "epoch": 0.10296561241445348,
+      "grad_norm": 0.5200375326910797,
+      "learning_rate": 4.961767366498055e-06,
+      "loss": 0.201,
+      "step": 457
+    },
+    {
+      "epoch": 0.10319092010026192,
+      "grad_norm": 0.5145908578521863,
+      "learning_rate": 4.961442130849933e-06,
+      "loss": 0.194,
+      "step": 458
+    },
+    {
+      "epoch": 0.10341622778607035,
+      "grad_norm": 0.518272482827854,
+      "learning_rate": 4.961115528454743e-06,
+      "loss": 0.2125,
+      "step": 459
+    },
+    {
+      "epoch": 0.10364153547187878,
+      "grad_norm": 0.4877010421344951,
+      "learning_rate": 4.960787559493836e-06,
+      "loss": 0.1833,
+      "step": 460
+    },
+    {
+      "epoch": 0.10386684315768721,
+      "grad_norm": 0.5327688445453027,
+      "learning_rate": 4.96045822414932e-06,
+      "loss": 0.1971,
+      "step": 461
+    },
+    {
+      "epoch": 0.10409215084349564,
+      "grad_norm": 0.5036310460862861,
+      "learning_rate": 4.960127522604065e-06,
+      "loss": 0.1895,
+      "step": 462
+    },
+    {
+      "epoch": 0.10431745852930407,
+      "grad_norm": 0.5025301179355007,
+      "learning_rate": 4.959795455041694e-06,
+      "loss": 0.1883,
+      "step": 463
+    },
+    {
+      "epoch": 0.1045427662151125,
+      "grad_norm": 0.5472446648835756,
+      "learning_rate": 4.959462021646593e-06,
+      "loss": 0.201,
+      "step": 464
+    },
+    {
+      "epoch": 0.10476807390092094,
+      "grad_norm": 0.4730210205377272,
+      "learning_rate": 4.959127222603905e-06,
+      "loss": 0.1696,
+      "step": 465
+    },
+    {
+      "epoch": 0.10499338158672938,
+      "grad_norm": 0.5300199033723536,
+      "learning_rate": 4.958791058099533e-06,
+      "loss": 0.2005,
+      "step": 466
+    },
+    {
+      "epoch": 0.10521868927253782,
+      "grad_norm": 0.49556193439744467,
+      "learning_rate": 4.958453528320135e-06,
+      "loss": 0.1804,
+      "step": 467
+    },
+    {
+      "epoch": 0.10544399695834625,
+      "grad_norm": 0.4865416664678349,
+      "learning_rate": 4.95811463345313e-06,
+      "loss": 0.1928,
+      "step": 468
+    },
+    {
+      "epoch": 0.10566930464415468,
+      "grad_norm": 0.5363619996168147,
+      "learning_rate": 4.957774373686692e-06,
+      "loss": 0.21,
+      "step": 469
+    },
+    {
+      "epoch": 0.10589461232996311,
+      "grad_norm": 0.5064762335561315,
+      "learning_rate": 4.957432749209755e-06,
+      "loss": 0.2008,
+      "step": 470
+    },
+    {
+      "epoch": 0.10611992001577154,
+      "grad_norm": 0.5144385896465787,
+      "learning_rate": 4.95708976021201e-06,
+      "loss": 0.1943,
+      "step": 471
+    },
+    {
+      "epoch": 0.10634522770157998,
+      "grad_norm": 0.4726824280116551,
+      "learning_rate": 4.956745406883909e-06,
+      "loss": 0.1933,
+      "step": 472
+    },
+    {
+      "epoch": 0.10657053538738841,
+      "grad_norm": 0.47149073477224396,
+      "learning_rate": 4.956399689416654e-06,
+      "loss": 0.1814,
+      "step": 473
+    },
+    {
+      "epoch": 0.10679584307319684,
+      "grad_norm": 0.5120111770818855,
+      "learning_rate": 4.956052608002212e-06,
+      "loss": 0.1997,
+      "step": 474
+    },
+    {
+      "epoch": 0.10702115075900527,
+      "grad_norm": 0.5031173640502962,
+      "learning_rate": 4.9557041628333046e-06,
+      "loss": 0.1989,
+      "step": 475
+    },
+    {
+      "epoch": 0.1072464584448137,
+      "grad_norm": 0.5063321024933504,
+      "learning_rate": 4.9553543541034086e-06,
+      "loss": 0.1733,
+      "step": 476
+    },
+    {
+      "epoch": 0.10747176613062213,
+      "grad_norm": 0.4792989721434702,
+      "learning_rate": 4.955003182006761e-06,
+      "loss": 0.1894,
+      "step": 477
+    },
+    {
+      "epoch": 0.10769707381643057,
+      "grad_norm": 0.5184160979761754,
+      "learning_rate": 4.954650646738354e-06,
+      "loss": 0.1981,
+      "step": 478
+    },
+    {
+      "epoch": 0.107922381502239,
+      "grad_norm": 0.4994948610360651,
+      "learning_rate": 4.954296748493938e-06,
+      "loss": 0.1903,
+      "step": 479
+    },
+    {
+      "epoch": 0.10814768918804743,
+      "grad_norm": 0.49550630583633326,
+      "learning_rate": 4.953941487470017e-06,
+      "loss": 0.1872,
+      "step": 480
+    },
+    {
+      "epoch": 0.10837299687385586,
+      "grad_norm": 0.4979423138808952,
+      "learning_rate": 4.9535848638638586e-06,
+      "loss": 0.1916,
+      "step": 481
+    },
+    {
+      "epoch": 0.1085983045596643,
+      "grad_norm": 0.5108535093532186,
+      "learning_rate": 4.953226877873479e-06,
+      "loss": 0.2137,
+      "step": 482
+    },
+    {
+      "epoch": 0.10882361224547273,
+      "grad_norm": 0.4914789380535188,
+      "learning_rate": 4.952867529697656e-06,
+      "loss": 0.1898,
+      "step": 483
+    },
+    {
+      "epoch": 0.10904891993128116,
+      "grad_norm": 0.4996741383080328,
+      "learning_rate": 4.952506819535922e-06,
+      "loss": 0.193,
+      "step": 484
+    },
+    {
+      "epoch": 0.10927422761708959,
+      "grad_norm": 0.4756841066506028,
+      "learning_rate": 4.952144747588566e-06,
+      "loss": 0.2049,
+      "step": 485
+    },
+    {
+      "epoch": 0.10949953530289802,
+      "grad_norm": 0.5312558416236551,
+      "learning_rate": 4.951781314056633e-06,
+      "loss": 0.196,
+      "step": 486
+    },
+    {
+      "epoch": 0.10972484298870645,
+      "grad_norm": 0.47800268646991784,
+      "learning_rate": 4.951416519141923e-06,
+      "loss": 0.1961,
+      "step": 487
+    },
+    {
+      "epoch": 0.10995015067451488,
+      "grad_norm": 0.4866331740965868,
+      "learning_rate": 4.951050363046995e-06,
+      "loss": 0.1779,
+      "step": 488
+    },
+    {
+      "epoch": 0.11017545836032332,
+      "grad_norm": 0.5149337838452025,
+      "learning_rate": 4.95068284597516e-06,
+      "loss": 0.2117,
+      "step": 489
+    },
+    {
+      "epoch": 0.11040076604613175,
+      "grad_norm": 0.5196237692139258,
+      "learning_rate": 4.950313968130488e-06,
+      "loss": 0.2082,
+      "step": 490
+    },
+    {
+      "epoch": 0.11062607373194018,
+      "grad_norm": 0.5041250602592621,
+      "learning_rate": 4.949943729717802e-06,
+      "loss": 0.1971,
+      "step": 491
+    },
+    {
+      "epoch": 0.11085138141774861,
+      "grad_norm": 0.476324372109053,
+      "learning_rate": 4.949572130942683e-06,
+      "loss": 0.1911,
+      "step": 492
+    },
+    {
+      "epoch": 0.11107668910355704,
+      "grad_norm": 0.46691888238357065,
+      "learning_rate": 4.949199172011464e-06,
+      "loss": 0.1881,
+      "step": 493
+    },
+    {
+      "epoch": 0.11130199678936548,
+      "grad_norm": 0.5370756908861958,
+      "learning_rate": 4.948824853131237e-06,
+      "loss": 0.2091,
+      "step": 494
+    },
+    {
+      "epoch": 0.11152730447517391,
+      "grad_norm": 0.481306717114356,
+      "learning_rate": 4.948449174509846e-06,
+      "loss": 0.1981,
+      "step": 495
+    },
+    {
+      "epoch": 0.11175261216098234,
+      "grad_norm": 0.5240621763320067,
+      "learning_rate": 4.948072136355892e-06,
+      "loss": 0.1869,
+      "step": 496
+    },
+    {
+      "epoch": 0.11197791984679077,
+      "grad_norm": 0.5353721219133655,
+      "learning_rate": 4.94769373887873e-06,
+      "loss": 0.2026,
+      "step": 497
+    },
+    {
+      "epoch": 0.1122032275325992,
+      "grad_norm": 0.48135528686973533,
+      "learning_rate": 4.94731398228847e-06,
+      "loss": 0.2066,
+      "step": 498
+    },
+    {
+      "epoch": 0.11242853521840764,
+      "grad_norm": 0.5284129479374139,
+      "learning_rate": 4.946932866795977e-06,
+      "loss": 0.1989,
+      "step": 499
+    },
+    {
+      "epoch": 0.11265384290421607,
+      "grad_norm": 0.5209491252410734,
+      "learning_rate": 4.94655039261287e-06,
+      "loss": 0.1991,
+      "step": 500
+    },
+    {
+      "epoch": 0.11265384290421607,
+      "eval_loss": 0.19234976172447205,
+      "eval_runtime": 56.8076,
+      "eval_samples_per_second": 50.521,
+      "eval_steps_per_second": 6.32,
+      "step": 500
+    },
+    {
+      "epoch": 0.1128791505900245,
+      "grad_norm": 0.4663315177005309,
+      "learning_rate": 4.946166559951523e-06,
+      "loss": 0.1823,
+      "step": 501
+    },
+    {
+      "epoch": 0.11310445827583293,
+      "grad_norm": 0.5043217933291648,
+      "learning_rate": 4.9457813690250635e-06,
+      "loss": 0.1964,
+      "step": 502
+    },
+    {
+      "epoch": 0.11332976596164136,
+      "grad_norm": 0.5055573412209956,
+      "learning_rate": 4.945394820047373e-06,
+      "loss": 0.1881,
+      "step": 503
+    },
+    {
+      "epoch": 0.1135550736474498,
+      "grad_norm": 0.48836811641356404,
+      "learning_rate": 4.94500691323309e-06,
+      "loss": 0.1848,
+      "step": 504
+    },
+    {
+      "epoch": 0.11378038133325823,
+      "grad_norm": 0.5455722455051317,
+      "learning_rate": 4.944617648797602e-06,
+      "loss": 0.2136,
+      "step": 505
+    },
+    {
+      "epoch": 0.11400568901906666,
+      "grad_norm": 0.4926313400228012,
+      "learning_rate": 4.9442270269570545e-06,
+      "loss": 0.1751,
+      "step": 506
+    },
+    {
+      "epoch": 0.11423099670487509,
+      "grad_norm": 0.4915724081406104,
+      "learning_rate": 4.943835047928346e-06,
+      "loss": 0.1686,
+      "step": 507
+    },
+    {
+      "epoch": 0.11445630439068352,
+      "grad_norm": 0.5428277431714841,
+      "learning_rate": 4.943441711929126e-06,
+      "loss": 0.2063,
+      "step": 508
+    },
+    {
+      "epoch": 0.11468161207649195,
+      "grad_norm": 0.5182829108034406,
+      "learning_rate": 4.9430470191778e-06,
+      "loss": 0.2048,
+      "step": 509
+    },
+    {
+      "epoch": 0.11490691976230039,
+      "grad_norm": 0.4919127976132606,
+      "learning_rate": 4.942650969893527e-06,
+      "loss": 0.1903,
+      "step": 510
+    },
+    {
+      "epoch": 0.11513222744810882,
+      "grad_norm": 0.5136285948187803,
+      "learning_rate": 4.942253564296217e-06,
+      "loss": 0.2014,
+      "step": 511
+    },
+    {
+      "epoch": 0.11535753513391725,
+      "grad_norm": 0.48212990917389625,
+      "learning_rate": 4.941854802606537e-06,
+      "loss": 0.1882,
+      "step": 512
+    },
+    {
+      "epoch": 0.11558284281972568,
+      "grad_norm": 0.4586425822677684,
+      "learning_rate": 4.9414546850459014e-06,
+      "loss": 0.175,
+      "step": 513
+    },
+    {
+      "epoch": 0.11580815050553411,
+      "grad_norm": 0.47705894227746765,
+      "learning_rate": 4.941053211836482e-06,
+      "loss": 0.1764,
+      "step": 514
+    },
+    {
+      "epoch": 0.11603345819134256,
+      "grad_norm": 0.5167055192881297,
+      "learning_rate": 4.940650383201202e-06,
+      "loss": 0.1909,
+      "step": 515
+    },
+    {
+      "epoch": 0.11625876587715099,
+      "grad_norm": 0.5060498826515504,
+      "learning_rate": 4.940246199363737e-06,
+      "loss": 0.1867,
+      "step": 516
+    },
+    {
+      "epoch": 0.11648407356295942,
+      "grad_norm": 0.5104025646241316,
+      "learning_rate": 4.939840660548515e-06,
+      "loss": 0.1956,
+      "step": 517
+    },
+    {
+      "epoch": 0.11670938124876785,
+      "grad_norm": 0.505292042282106,
+      "learning_rate": 4.939433766980717e-06,
+      "loss": 0.1794,
+      "step": 518
+    },
+    {
+      "epoch": 0.11693468893457629,
+      "grad_norm": 0.5223802995592578,
+      "learning_rate": 4.939025518886276e-06,
+      "loss": 0.1986,
+      "step": 519
+    },
+    {
+      "epoch": 0.11715999662038472,
+      "grad_norm": 0.47521330675412077,
+      "learning_rate": 4.9386159164918764e-06,
+      "loss": 0.183,
+      "step": 520
+    },
+    {
+      "epoch": 0.11738530430619315,
+      "grad_norm": 0.5750076163542128,
+      "learning_rate": 4.938204960024955e-06,
+      "loss": 0.2033,
+      "step": 521
+    },
+    {
+      "epoch": 0.11761061199200158,
+      "grad_norm": 0.5233779578600689,
+      "learning_rate": 4.937792649713701e-06,
+      "loss": 0.1864,
+      "step": 522
+    },
+    {
+      "epoch": 0.11783591967781001,
+      "grad_norm": 0.46217847479401275,
+      "learning_rate": 4.937378985787055e-06,
+      "loss": 0.1753,
+      "step": 523
+    },
+    {
+      "epoch": 0.11806122736361845,
+      "grad_norm": 0.4974554874716037,
+      "learning_rate": 4.9369639684747095e-06,
+      "loss": 0.1836,
+      "step": 524
+    },
+    {
+      "epoch": 0.11828653504942688,
+      "grad_norm": 0.4923126383015336,
+      "learning_rate": 4.936547598007107e-06,
+      "loss": 0.193,
+      "step": 525
+    },
+    {
+      "epoch": 0.11851184273523531,
+      "grad_norm": 0.5398798000068692,
+      "learning_rate": 4.936129874615443e-06,
+      "loss": 0.1755,
+      "step": 526
+    },
+    {
+      "epoch": 0.11873715042104374,
+      "grad_norm": 0.47146590078183515,
+      "learning_rate": 4.935710798531664e-06,
+      "loss": 0.1971,
+      "step": 527
+    },
+    {
+      "epoch": 0.11896245810685217,
+      "grad_norm": 0.5491858771837667,
+      "learning_rate": 4.935290369988468e-06,
+      "loss": 0.2056,
+      "step": 528
+    },
+    {
+      "epoch": 0.1191877657926606,
+      "grad_norm": 0.5388147308207761,
+      "learning_rate": 4.934868589219302e-06,
+      "loss": 0.2001,
+      "step": 529
+    },
+    {
+      "epoch": 0.11941307347846904,
+      "grad_norm": 0.5083952861725305,
+      "learning_rate": 4.934445456458366e-06,
+      "loss": 0.1888,
+      "step": 530
+    },
+    {
+      "epoch": 0.11963838116427747,
+      "grad_norm": 0.5169297785302316,
+      "learning_rate": 4.934020971940609e-06,
+      "loss": 0.2023,
+      "step": 531
+    },
+    {
+      "epoch": 0.1198636888500859,
+      "grad_norm": 0.5095162486691678,
+      "learning_rate": 4.933595135901733e-06,
+      "loss": 0.1906,
+      "step": 532
+    },
+    {
+      "epoch": 0.12008899653589433,
+      "grad_norm": 0.48223521419999554,
+      "learning_rate": 4.933167948578187e-06,
+      "loss": 0.1838,
+      "step": 533
+    },
+    {
+      "epoch": 0.12031430422170276,
+      "grad_norm": 0.5444055294135087,
+      "learning_rate": 4.932739410207172e-06,
+      "loss": 0.2002,
+      "step": 534
+    },
+    {
+      "epoch": 0.1205396119075112,
+      "grad_norm": 0.5109454104003555,
+      "learning_rate": 4.932309521026643e-06,
+      "loss": 0.1927,
+      "step": 535
+    },
+    {
+      "epoch": 0.12076491959331963,
+      "grad_norm": 0.48724028911776623,
+      "learning_rate": 4.931878281275296e-06,
+      "loss": 0.1844,
+      "step": 536
+    },
+    {
+      "epoch": 0.12099022727912806,
+      "grad_norm": 0.4777988104796984,
+      "learning_rate": 4.931445691192587e-06,
+      "loss": 0.1785,
+      "step": 537
+    },
+    {
+      "epoch": 0.12121553496493649,
+      "grad_norm": 0.5021184008826821,
+      "learning_rate": 4.931011751018715e-06,
+      "loss": 0.1883,
+      "step": 538
+    },
+    {
+      "epoch": 0.12144084265074492,
+      "grad_norm": 0.45767088453705446,
+      "learning_rate": 4.930576460994631e-06,
+      "loss": 0.1744,
+      "step": 539
+    },
+    {
+      "epoch": 0.12166615033655336,
+      "grad_norm": 0.5156528768504517,
+      "learning_rate": 4.930139821362036e-06,
+      "loss": 0.1957,
+      "step": 540
+    },
+    {
+      "epoch": 0.12189145802236179,
+      "grad_norm": 0.49466412609385285,
+      "learning_rate": 4.929701832363379e-06,
+      "loss": 0.1932,
+      "step": 541
+    },
+    {
+      "epoch": 0.12211676570817022,
+      "grad_norm": 0.454600799745477,
+      "learning_rate": 4.929262494241859e-06,
+      "loss": 0.173,
+      "step": 542
+    },
+    {
+      "epoch": 0.12234207339397865,
+      "grad_norm": 0.48395076267227993,
+      "learning_rate": 4.928821807241425e-06,
+      "loss": 0.1864,
+      "step": 543
+    },
+    {
+      "epoch": 0.12256738107978708,
+      "grad_norm": 0.472630796102665,
+      "learning_rate": 4.928379771606773e-06,
+      "loss": 0.1827,
+      "step": 544
+    },
+    {
+      "epoch": 0.12279268876559551,
+      "grad_norm": 0.5131023372757877,
+      "learning_rate": 4.927936387583348e-06,
+      "loss": 0.2072,
+      "step": 545
+    },
+    {
+      "epoch": 0.12301799645140395,
+      "grad_norm": 0.4599619189881397,
+      "learning_rate": 4.927491655417347e-06,
+      "loss": 0.1772,
+      "step": 546
+    },
+    {
+      "epoch": 0.12324330413721238,
+      "grad_norm": 0.4936735496643357,
+      "learning_rate": 4.927045575355712e-06,
+      "loss": 0.2063,
+      "step": 547
+    },
+    {
+      "epoch": 0.12346861182302081,
+      "grad_norm": 0.5223093419687126,
+      "learning_rate": 4.926598147646134e-06,
+      "loss": 0.1932,
+      "step": 548
+    },
+    {
+      "epoch": 0.12369391950882924,
+      "grad_norm": 0.458455620184728,
+      "learning_rate": 4.9261493725370546e-06,
+      "loss": 0.1677,
+      "step": 549
+    },
+    {
+      "epoch": 0.12391922719463767,
+      "grad_norm": 0.49240690720846386,
+      "learning_rate": 4.9256992502776605e-06,
+      "loss": 0.1839,
+      "step": 550
+    },
+    {
+      "epoch": 0.1241445348804461,
+      "grad_norm": 0.47757607892838266,
+      "learning_rate": 4.925247781117888e-06,
+      "loss": 0.1733,
+      "step": 551
+    },
+    {
+      "epoch": 0.12436984256625454,
+      "grad_norm": 0.4838343579732949,
+      "learning_rate": 4.924794965308421e-06,
+      "loss": 0.1763,
+      "step": 552
+    },
+    {
+      "epoch": 0.12459515025206297,
+      "grad_norm": 0.5265568200531762,
+      "learning_rate": 4.924340803100692e-06,
+      "loss": 0.2014,
+      "step": 553
+    },
+    {
+      "epoch": 0.1248204579378714,
+      "grad_norm": 0.48214538271531815,
+      "learning_rate": 4.9238852947468796e-06,
+      "loss": 0.1857,
+      "step": 554
+    },
+    {
+      "epoch": 0.12504576562367983,
+      "grad_norm": 0.5343120785647222,
+      "learning_rate": 4.923428440499912e-06,
+      "loss": 0.1906,
+      "step": 555
+    },
+    {
+      "epoch": 0.12527107330948828,
+      "grad_norm": 0.5235638112156517,
+      "learning_rate": 4.922970240613461e-06,
+      "loss": 0.191,
+      "step": 556
+    },
+    {
+      "epoch": 0.1254963809952967,
+      "grad_norm": 0.4889226147171265,
+      "learning_rate": 4.92251069534195e-06,
+      "loss": 0.1949,
+      "step": 557
+    },
+    {
+      "epoch": 0.12572168868110514,
+      "grad_norm": 0.5040294823792546,
+      "learning_rate": 4.922049804940546e-06,
+      "loss": 0.194,
+      "step": 558
+    },
+    {
+      "epoch": 0.12594699636691356,
+      "grad_norm": 0.5391772652484352,
+      "learning_rate": 4.9215875696651645e-06,
+      "loss": 0.1944,
+      "step": 559
+    },
+    {
+      "epoch": 0.126172304052722,
+      "grad_norm": 0.5025810534944168,
+      "learning_rate": 4.9211239897724685e-06,
+      "loss": 0.1982,
+      "step": 560
+    },
+    {
+      "epoch": 0.12639761173853042,
+      "grad_norm": 0.4798927968328912,
+      "learning_rate": 4.920659065519866e-06,
+      "loss": 0.1968,
+      "step": 561
+    },
+    {
+      "epoch": 0.12662291942433887,
+      "grad_norm": 0.5166664453113821,
+      "learning_rate": 4.920192797165511e-06,
+      "loss": 0.1843,
+      "step": 562
+    },
+    {
+      "epoch": 0.1268482271101473,
+      "grad_norm": 0.567222736069852,
+      "learning_rate": 4.919725184968307e-06,
+      "loss": 0.1982,
+      "step": 563
+    },
+    {
+      "epoch": 0.12707353479595573,
+      "grad_norm": 0.48741393189065063,
+      "learning_rate": 4.9192562291879e-06,
+      "loss": 0.184,
+      "step": 564
+    },
+    {
+      "epoch": 0.12729884248176415,
+      "grad_norm": 0.5093830078546595,
+      "learning_rate": 4.9187859300846845e-06,
+      "loss": 0.1852,
+      "step": 565
+    },
+    {
+      "epoch": 0.1275241501675726,
+      "grad_norm": 0.4947704192884275,
+      "learning_rate": 4.9183142879198e-06,
+      "loss": 0.1991,
+      "step": 566
+    },
+    {
+      "epoch": 0.12774945785338102,
+      "grad_norm": 0.48047644785859844,
+      "learning_rate": 4.917841302955132e-06,
+      "loss": 0.189,
+      "step": 567
+    },
+    {
+      "epoch": 0.12797476553918946,
+      "grad_norm": 0.49149253370292334,
+      "learning_rate": 4.917366975453311e-06,
+      "loss": 0.1846,
+      "step": 568
+    },
+    {
+      "epoch": 0.12820007322499788,
+      "grad_norm": 0.4772556842972461,
+      "learning_rate": 4.916891305677712e-06,
+      "loss": 0.1812,
+      "step": 569
+    },
+    {
+      "epoch": 0.12842538091080632,
+      "grad_norm": 0.4967161370998577,
+      "learning_rate": 4.9164142938924595e-06,
+      "loss": 0.1943,
+      "step": 570
+    },
+    {
+      "epoch": 0.12865068859661474,
+      "grad_norm": 0.4966530996832579,
+      "learning_rate": 4.9159359403624185e-06,
+      "loss": 0.1963,
+      "step": 571
+    },
+    {
+      "epoch": 0.1288759962824232,
+      "grad_norm": 0.49264929970143695,
+      "learning_rate": 4.915456245353202e-06,
+      "loss": 0.1752,
+      "step": 572
+    },
+    {
+      "epoch": 0.1291013039682316,
+      "grad_norm": 0.4859626874941514,
+      "learning_rate": 4.914975209131165e-06,
+      "loss": 0.1937,
+      "step": 573
+    },
+    {
+      "epoch": 0.12932661165404005,
+      "grad_norm": 0.5043817672029466,
+      "learning_rate": 4.914492831963411e-06,
+      "loss": 0.1937,
+      "step": 574
+    },
+    {
+      "epoch": 0.12955191933984847,
+      "grad_norm": 0.5015582050349247,
+      "learning_rate": 4.9140091141177856e-06,
+      "loss": 0.201,
+      "step": 575
+    },
+    {
+      "epoch": 0.12977722702565692,
+      "grad_norm": 0.5146518830297795,
+      "learning_rate": 4.9135240558628786e-06,
+      "loss": 0.2031,
+      "step": 576
+    },
+    {
+      "epoch": 0.13000253471146533,
+      "grad_norm": 0.4854388066659389,
+      "learning_rate": 4.913037657468025e-06,
+      "loss": 0.1913,
+      "step": 577
+    },
+    {
+      "epoch": 0.13022784239727378,
+      "grad_norm": 0.48652005999572784,
+      "learning_rate": 4.9125499192033035e-06,
+      "loss": 0.1823,
+      "step": 578
+    },
+    {
+      "epoch": 0.1304531500830822,
+      "grad_norm": 0.46886496437160863,
+      "learning_rate": 4.912060841339536e-06,
+      "loss": 0.1681,
+      "step": 579
+    },
+    {
+      "epoch": 0.13067845776889064,
+      "grad_norm": 0.4733097967338571,
+      "learning_rate": 4.911570424148293e-06,
+      "loss": 0.1781,
+      "step": 580
+    },
+    {
+      "epoch": 0.13090376545469906,
+      "grad_norm": 0.4966002182337164,
+      "learning_rate": 4.911078667901881e-06,
+      "loss": 0.2003,
+      "step": 581
+    },
+    {
+      "epoch": 0.1311290731405075,
+      "grad_norm": 0.46547777291654663,
+      "learning_rate": 4.910585572873355e-06,
+      "loss": 0.1814,
+      "step": 582
+    },
+    {
+      "epoch": 0.13135438082631593,
+      "grad_norm": 0.500486589164084,
+      "learning_rate": 4.9100911393365134e-06,
+      "loss": 0.1627,
+      "step": 583
+    },
+    {
+      "epoch": 0.13157968851212437,
+      "grad_norm": 0.4955517290997085,
+      "learning_rate": 4.9095953675658945e-06,
+      "loss": 0.1972,
+      "step": 584
+    },
+    {
+      "epoch": 0.1318049961979328,
+      "grad_norm": 0.4988068045683488,
+      "learning_rate": 4.909098257836784e-06,
+      "loss": 0.1696,
+      "step": 585
+    },
+    {
+      "epoch": 0.13203030388374123,
+      "grad_norm": 0.5068279577450977,
+      "learning_rate": 4.908599810425208e-06,
+      "loss": 0.1959,
+      "step": 586
+    },
+    {
+      "epoch": 0.13225561156954965,
+      "grad_norm": 0.4847712941015659,
+      "learning_rate": 4.908100025607935e-06,
+      "loss": 0.1839,
+      "step": 587
+    },
+    {
+      "epoch": 0.1324809192553581,
+      "grad_norm": 0.4745230551209773,
+      "learning_rate": 4.907598903662477e-06,
+      "loss": 0.1774,
+      "step": 588
+    },
+    {
+      "epoch": 0.13270622694116654,
+      "grad_norm": 0.5141740196225114,
+      "learning_rate": 4.90709644486709e-06,
+      "loss": 0.2046,
+      "step": 589
+    },
+    {
+      "epoch": 0.13293153462697496,
+      "grad_norm": 0.5221614658472497,
+      "learning_rate": 4.906592649500767e-06,
+      "loss": 0.1908,
+      "step": 590
+    },
+    {
+      "epoch": 0.1331568423127834,
+      "grad_norm": 0.4940345638894506,
+      "learning_rate": 4.906087517843251e-06,
+      "loss": 0.1838,
+      "step": 591
+    },
+    {
+      "epoch": 0.13338214999859183,
+      "grad_norm": 0.4688033247262004,
+      "learning_rate": 4.9055810501750205e-06,
+      "loss": 0.1766,
+      "step": 592
+    },
+    {
+      "epoch": 0.13360745768440027,
+      "grad_norm": 0.48052014823475525,
+      "learning_rate": 4.905073246777298e-06,
+      "loss": 0.1848,
+      "step": 593
+    },
+    {
+      "epoch": 0.1338327653702087,
+      "grad_norm": 0.48977849616782837,
+      "learning_rate": 4.904564107932048e-06,
+      "loss": 0.1737,
+      "step": 594
+    },
+    {
+      "epoch": 0.13405807305601714,
+      "grad_norm": 0.49857557136653785,
+      "learning_rate": 4.904053633921977e-06,
+      "loss": 0.1812,
+      "step": 595
+    },
+    {
+      "epoch": 0.13428338074182555,
+      "grad_norm": 0.5195821127291367,
+      "learning_rate": 4.9035418250305314e-06,
+      "loss": 0.1994,
+      "step": 596
+    },
+    {
+      "epoch": 0.134508688427634,
+      "grad_norm": 0.4848753612469132,
+      "learning_rate": 4.9030286815419e-06,
+      "loss": 0.1746,
+      "step": 597
+    },
+    {
+      "epoch": 0.13473399611344242,
+      "grad_norm": 0.4651188815320874,
+      "learning_rate": 4.902514203741013e-06,
+      "loss": 0.1717,
+      "step": 598
+    },
+    {
+      "epoch": 0.13495930379925086,
+      "grad_norm": 0.49092504275491067,
+      "learning_rate": 4.901998391913539e-06,
+      "loss": 0.1742,
+      "step": 599
+    },
+    {
+      "epoch": 0.13518461148505928,
+      "grad_norm": 0.48578109490095983,
+      "learning_rate": 4.9014812463458905e-06,
+      "loss": 0.1768,
+      "step": 600
+    },
+    {
+      "epoch": 0.13540991917086773,
+      "grad_norm": 0.5097924287582863,
+      "learning_rate": 4.9009627673252195e-06,
+      "loss": 0.1925,
+      "step": 601
+    },
+    {
+      "epoch": 0.13563522685667614,
+      "grad_norm": 0.49925568640506196,
+      "learning_rate": 4.9004429551394155e-06,
+      "loss": 0.1901,
+      "step": 602
+    },
+    {
+      "epoch": 0.1358605345424846,
+      "grad_norm": 0.5160976182113821,
+      "learning_rate": 4.899921810077114e-06,
+      "loss": 0.194,
+      "step": 603
+    },
+    {
+      "epoch": 0.136085842228293,
+      "grad_norm": 0.5130419470354679,
+      "learning_rate": 4.899399332427685e-06,
+      "loss": 0.181,
+      "step": 604
+    },
+    {
+      "epoch": 0.13631114991410145,
+      "grad_norm": 0.49486082839466233,
+      "learning_rate": 4.898875522481242e-06,
+      "loss": 0.1853,
+      "step": 605
+    },
+    {
+      "epoch": 0.13653645759990987,
+      "grad_norm": 0.4648043819729389,
+      "learning_rate": 4.898350380528638e-06,
+      "loss": 0.1893,
+      "step": 606
+    },
+    {
+      "epoch": 0.13676176528571832,
+      "grad_norm": 0.4743562941887421,
+      "learning_rate": 4.897823906861463e-06,
+      "loss": 0.1919,
+      "step": 607
+    },
+    {
+      "epoch": 0.13698707297152674,
+      "grad_norm": 0.49175914667017384,
+      "learning_rate": 4.89729610177205e-06,
+      "loss": 0.1876,
+      "step": 608
+    },
+    {
+      "epoch": 0.13721238065733518,
+      "grad_norm": 0.5060333088563503,
+      "learning_rate": 4.896766965553467e-06,
+      "loss": 0.208,
+      "step": 609
+    },
+    {
+      "epoch": 0.1374376883431436,
+      "grad_norm": 0.47294188413884825,
+      "learning_rate": 4.896236498499526e-06,
+      "loss": 0.1988,
+      "step": 610
+    },
+    {
+      "epoch": 0.13766299602895204,
+      "grad_norm": 0.4606626602858298,
+      "learning_rate": 4.8957047009047744e-06,
+      "loss": 0.1752,
+      "step": 611
+    },
+    {
+      "epoch": 0.13788830371476046,
+      "grad_norm": 0.47273517792676306,
+      "learning_rate": 4.8951715730645e-06,
+      "loss": 0.1884,
+      "step": 612
+    },
+    {
+      "epoch": 0.1381136114005689,
+      "grad_norm": 0.46148619697171755,
+      "learning_rate": 4.894637115274728e-06,
+      "loss": 0.1885,
+      "step": 613
+    },
+    {
+      "epoch": 0.13833891908637733,
+      "grad_norm": 0.5018754666434547,
+      "learning_rate": 4.894101327832225e-06,
+      "loss": 0.2106,
+      "step": 614
+    },
+    {
+      "epoch": 0.13856422677218577,
+      "grad_norm": 0.4651687639877327,
+      "learning_rate": 4.893564211034492e-06,
+      "loss": 0.1792,
+      "step": 615
+    },
+    {
+      "epoch": 0.1387895344579942,
+      "grad_norm": 0.47670934711892976,
+      "learning_rate": 4.89302576517977e-06,
+      "loss": 0.1896,
+      "step": 616
+    },
+    {
+      "epoch": 0.13901484214380264,
+      "grad_norm": 0.4545326790236107,
+      "learning_rate": 4.892485990567037e-06,
+      "loss": 0.1833,
+      "step": 617
+    },
+    {
+      "epoch": 0.13924014982961105,
+      "grad_norm": 0.5051419942175441,
+      "learning_rate": 4.891944887496013e-06,
+      "loss": 0.2025,
+      "step": 618
+    },
+    {
+      "epoch": 0.1394654575154195,
+      "grad_norm": 0.511578758034126,
+      "learning_rate": 4.891402456267149e-06,
+      "loss": 0.1922,
+      "step": 619
+    },
+    {
+      "epoch": 0.13969076520122792,
+      "grad_norm": 0.4935796436295554,
+      "learning_rate": 4.890858697181638e-06,
+      "loss": 0.1915,
+      "step": 620
+    },
+    {
+      "epoch": 0.13991607288703636,
+      "grad_norm": 0.502114052015495,
+      "learning_rate": 4.89031361054141e-06,
+      "loss": 0.1922,
+      "step": 621
+    },
+    {
+      "epoch": 0.14014138057284478,
+      "grad_norm": 0.510065283262986,
+      "learning_rate": 4.8897671966491315e-06,
+      "loss": 0.1826,
+      "step": 622
+    },
+    {
+      "epoch": 0.14036668825865323,
+      "grad_norm": 0.4952918551445192,
+      "learning_rate": 4.889219455808204e-06,
+      "loss": 0.1854,
+      "step": 623
+    },
+    {
+      "epoch": 0.14059199594446165,
+      "grad_norm": 0.5052374937308469,
+      "learning_rate": 4.888670388322768e-06,
+      "loss": 0.1985,
+      "step": 624
+    },
+    {
+      "epoch": 0.1408173036302701,
+      "grad_norm": 0.49988160987866054,
+      "learning_rate": 4.888119994497701e-06,
+      "loss": 0.1749,
+      "step": 625
+    },
+    {
+      "epoch": 0.1410426113160785,
+      "grad_norm": 0.5172442914773804,
+      "learning_rate": 4.887568274638616e-06,
+      "loss": 0.2019,
+      "step": 626
+    },
+    {
+      "epoch": 0.14126791900188695,
+      "grad_norm": 0.5087009835749207,
+      "learning_rate": 4.887015229051861e-06,
+      "loss": 0.191,
+      "step": 627
+    },
+    {
+      "epoch": 0.14149322668769537,
+      "grad_norm": 0.4959246585025635,
+      "learning_rate": 4.886460858044524e-06,
+      "loss": 0.1857,
+      "step": 628
+    },
+    {
+      "epoch": 0.14171853437350382,
+      "grad_norm": 0.4723135163607331,
+      "learning_rate": 4.885905161924426e-06,
+      "loss": 0.1822,
+      "step": 629
+    },
+    {
+      "epoch": 0.14194384205931224,
+      "grad_norm": 0.502793877350106,
+      "learning_rate": 4.8853481410001225e-06,
+      "loss": 0.1806,
+      "step": 630
+    },
+    {
+      "epoch": 0.14216914974512068,
+      "grad_norm": 0.5082071597376282,
+      "learning_rate": 4.8847897955809085e-06,
+      "loss": 0.1923,
+      "step": 631
+    },
+    {
+      "epoch": 0.1423944574309291,
+      "grad_norm": 0.46157106301262735,
+      "learning_rate": 4.884230125976812e-06,
+      "loss": 0.1673,
+      "step": 632
+    },
+    {
+      "epoch": 0.14261976511673755,
+      "grad_norm": 0.4917291509510748,
+      "learning_rate": 4.8836691324985955e-06,
+      "loss": 0.1935,
+      "step": 633
+    },
+    {
+      "epoch": 0.14284507280254596,
+      "grad_norm": 0.4936790877793207,
+      "learning_rate": 4.883106815457758e-06,
+      "loss": 0.193,
+      "step": 634
+    },
+    {
+      "epoch": 0.1430703804883544,
+      "grad_norm": 0.5003405509113616,
+      "learning_rate": 4.882543175166535e-06,
+      "loss": 0.1792,
+      "step": 635
+    },
+    {
+      "epoch": 0.14329568817416283,
+      "grad_norm": 0.5285886608390846,
+      "learning_rate": 4.881978211937895e-06,
+      "loss": 0.1984,
+      "step": 636
+    },
+    {
+      "epoch": 0.14352099585997127,
+      "grad_norm": 0.49050879740196224,
+      "learning_rate": 4.8814119260855374e-06,
+      "loss": 0.1872,
+      "step": 637
+    },
+    {
+      "epoch": 0.14374630354577972,
+      "grad_norm": 0.48682747731833537,
+      "learning_rate": 4.8808443179239025e-06,
+      "loss": 0.1755,
+      "step": 638
+    },
+    {
+      "epoch": 0.14397161123158814,
+      "grad_norm": 0.4676177116452303,
+      "learning_rate": 4.880275387768162e-06,
+      "loss": 0.1748,
+      "step": 639
+    },
+    {
+      "epoch": 0.14419691891739658,
+      "grad_norm": 0.4951558991868934,
+      "learning_rate": 4.87970513593422e-06,
+      "loss": 0.1843,
+      "step": 640
+    },
+    {
+      "epoch": 0.144422226603205,
+      "grad_norm": 0.5213133487107826,
+      "learning_rate": 4.879133562738719e-06,
+      "loss": 0.2134,
+      "step": 641
+    },
+    {
+      "epoch": 0.14464753428901345,
+      "grad_norm": 0.46810947829384675,
+      "learning_rate": 4.878560668499029e-06,
+      "loss": 0.1927,
+      "step": 642
+    },
+    {
+      "epoch": 0.14487284197482186,
+      "grad_norm": 0.47590414235063444,
+      "learning_rate": 4.8779864535332585e-06,
+      "loss": 0.1743,
+      "step": 643
+    },
+    {
+      "epoch": 0.1450981496606303,
+      "grad_norm": 0.4995226249566914,
+      "learning_rate": 4.877410918160247e-06,
+      "loss": 0.1815,
+      "step": 644
+    },
+    {
+      "epoch": 0.14532345734643873,
+      "grad_norm": 0.5278580391531498,
+      "learning_rate": 4.876834062699569e-06,
+      "loss": 0.2039,
+      "step": 645
+    },
+    {
+      "epoch": 0.14554876503224717,
+      "grad_norm": 0.4843992433234831,
+      "learning_rate": 4.87625588747153e-06,
+      "loss": 0.1935,
+      "step": 646
+    },
+    {
+      "epoch": 0.1457740727180556,
+      "grad_norm": 0.4854790733499744,
+      "learning_rate": 4.875676392797169e-06,
+      "loss": 0.1816,
+      "step": 647
+    },
+    {
+      "epoch": 0.14599938040386404,
+      "grad_norm": 0.5496275694036449,
+      "learning_rate": 4.875095578998258e-06,
+      "loss": 0.2017,
+      "step": 648
+    },
+    {
+      "epoch": 0.14622468808967246,
+      "grad_norm": 0.47840998998543804,
+      "learning_rate": 4.874513446397301e-06,
+      "loss": 0.1718,
+      "step": 649
+    },
+    {
+      "epoch": 0.1464499957754809,
+      "grad_norm": 0.472469437986709,
+      "learning_rate": 4.873929995317535e-06,
+      "loss": 0.1813,
+      "step": 650
+    },
+    {
+      "epoch": 0.14667530346128932,
+      "grad_norm": 0.49186188067027603,
+      "learning_rate": 4.873345226082929e-06,
+      "loss": 0.1874,
+      "step": 651
+    },
+    {
+      "epoch": 0.14690061114709776,
+      "grad_norm": 0.47817285908812057,
+      "learning_rate": 4.872759139018183e-06,
+      "loss": 0.1849,
+      "step": 652
+    },
+    {
+      "epoch": 0.14712591883290618,
+      "grad_norm": 0.5128159653562537,
+      "learning_rate": 4.872171734448728e-06,
+      "loss": 0.2035,
+      "step": 653
+    },
+    {
+      "epoch": 0.14735122651871463,
+      "grad_norm": 0.4996885022639693,
+      "learning_rate": 4.87158301270073e-06,
+      "loss": 0.1796,
+      "step": 654
+    },
+    {
+      "epoch": 0.14757653420452305,
+      "grad_norm": 0.4504567590267657,
+      "learning_rate": 4.870992974101084e-06,
+      "loss": 0.1778,
+      "step": 655
+    },
+    {
+      "epoch": 0.1478018418903315,
+      "grad_norm": 0.5086172528062239,
+      "learning_rate": 4.870401618977415e-06,
+      "loss": 0.2074,
+      "step": 656
+    },
+    {
+      "epoch": 0.1480271495761399,
+      "grad_norm": 0.49588512436516297,
+      "learning_rate": 4.869808947658082e-06,
+      "loss": 0.1972,
+      "step": 657
+    },
+    {
+      "epoch": 0.14825245726194836,
+      "grad_norm": 0.4917383983984287,
+      "learning_rate": 4.869214960472172e-06,
+      "loss": 0.1919,
+      "step": 658
+    },
+    {
+      "epoch": 0.14847776494775677,
+      "grad_norm": 0.510112771596669,
+      "learning_rate": 4.868619657749505e-06,
+      "loss": 0.1939,
+      "step": 659
+    },
+    {
+      "epoch": 0.14870307263356522,
+      "grad_norm": 0.5446853659905414,
+      "learning_rate": 4.868023039820629e-06,
+      "loss": 0.1979,
+      "step": 660
+    },
+    {
+      "epoch": 0.14892838031937364,
+      "grad_norm": 0.47307794313317136,
+      "learning_rate": 4.867425107016826e-06,
+      "loss": 0.1736,
+      "step": 661
+    },
+    {
+      "epoch": 0.14915368800518208,
+      "grad_norm": 0.4845079148956392,
+      "learning_rate": 4.8668258596701035e-06,
+      "loss": 0.166,
+      "step": 662
+    },
+    {
+      "epoch": 0.1493789956909905,
+      "grad_norm": 0.47209894690292553,
+      "learning_rate": 4.866225298113203e-06,
+      "loss": 0.175,
+      "step": 663
+    },
+    {
+      "epoch": 0.14960430337679895,
+      "grad_norm": 0.5525456803515458,
+      "learning_rate": 4.865623422679593e-06,
+      "loss": 0.1965,
+      "step": 664
+    },
+    {
+      "epoch": 0.14982961106260737,
+      "grad_norm": 0.5276283864774272,
+      "learning_rate": 4.865020233703472e-06,
+      "loss": 0.1967,
+      "step": 665
+    },
+    {
+      "epoch": 0.1500549187484158,
+      "grad_norm": 0.5237611032703069,
+      "learning_rate": 4.864415731519769e-06,
+      "loss": 0.1847,
+      "step": 666
+    },
+    {
+      "epoch": 0.15028022643422423,
+      "grad_norm": 0.49191733964287326,
+      "learning_rate": 4.863809916464142e-06,
+      "loss": 0.1843,
+      "step": 667
+    },
+    {
+      "epoch": 0.15050553412003267,
+      "grad_norm": 0.5046503834610336,
+      "learning_rate": 4.8632027888729765e-06,
+      "loss": 0.1873,
+      "step": 668
+    },
+    {
+      "epoch": 0.1507308418058411,
+      "grad_norm": 0.4761077464697414,
+      "learning_rate": 4.862594349083389e-06,
+      "loss": 0.1659,
+      "step": 669
+    },
+    {
+      "epoch": 0.15095614949164954,
+      "grad_norm": 0.49773885573902543,
+      "learning_rate": 4.861984597433223e-06,
+      "loss": 0.1857,
+      "step": 670
+    },
+    {
+      "epoch": 0.15118145717745796,
+      "grad_norm": 0.4657062167693691,
+      "learning_rate": 4.861373534261049e-06,
+      "loss": 0.171,
+      "step": 671
+    },
+    {
+      "epoch": 0.1514067648632664,
+      "grad_norm": 0.525891884598973,
+      "learning_rate": 4.860761159906171e-06,
+      "loss": 0.1975,
+      "step": 672
+    },
+    {
+      "epoch": 0.15163207254907482,
+      "grad_norm": 0.4543247451924693,
+      "learning_rate": 4.8601474747086145e-06,
+      "loss": 0.1705,
+      "step": 673
+    },
+    {
+      "epoch": 0.15185738023488327,
+      "grad_norm": 0.49799202938449977,
+      "learning_rate": 4.859532479009138e-06,
+      "loss": 0.192,
+      "step": 674
+    },
+    {
+      "epoch": 0.15208268792069168,
+      "grad_norm": 0.48331962531733624,
+      "learning_rate": 4.8589161731492255e-06,
+      "loss": 0.1753,
+      "step": 675
+    },
+    {
+      "epoch": 0.15230799560650013,
+      "grad_norm": 0.5156829087990366,
+      "learning_rate": 4.858298557471089e-06,
+      "loss": 0.1893,
+      "step": 676
+    },
+    {
+      "epoch": 0.15253330329230855,
+      "grad_norm": 0.5170730850666883,
+      "learning_rate": 4.857679632317664e-06,
+      "loss": 0.1965,
+      "step": 677
+    },
+    {
+      "epoch": 0.152758610978117,
+      "grad_norm": 0.5072378427728541,
+      "learning_rate": 4.857059398032622e-06,
+      "loss": 0.1816,
+      "step": 678
+    },
+    {
+      "epoch": 0.1529839186639254,
+      "grad_norm": 0.4898983755820026,
+      "learning_rate": 4.856437854960352e-06,
+      "loss": 0.1849,
+      "step": 679
+    },
+    {
+      "epoch": 0.15320922634973386,
+      "grad_norm": 0.5264472680458602,
+      "learning_rate": 4.855815003445975e-06,
+      "loss": 0.1971,
+      "step": 680
+    },
+    {
+      "epoch": 0.15343453403554227,
+      "grad_norm": 0.5152650134807022,
+      "learning_rate": 4.855190843835338e-06,
+      "loss": 0.1893,
+      "step": 681
+    },
+    {
+      "epoch": 0.15365984172135072,
+      "grad_norm": 0.5578264199187081,
+      "learning_rate": 4.8545653764750125e-06,
+      "loss": 0.2006,
+      "step": 682
+    },
+    {
+      "epoch": 0.15388514940715914,
+      "grad_norm": 0.46453655362572005,
+      "learning_rate": 4.853938601712297e-06,
+      "loss": 0.1805,
+      "step": 683
+    },
+    {
+      "epoch": 0.15411045709296758,
+      "grad_norm": 0.5213029966256437,
+      "learning_rate": 4.853310519895217e-06,
+      "loss": 0.1858,
+      "step": 684
+    },
+    {
+      "epoch": 0.154335764778776,
+      "grad_norm": 0.4915514653572081,
+      "learning_rate": 4.852681131372522e-06,
+      "loss": 0.1969,
+      "step": 685
+    },
+    {
+      "epoch": 0.15456107246458445,
+      "grad_norm": 0.5281210443216867,
+      "learning_rate": 4.85205043649369e-06,
+      "loss": 0.1912,
+      "step": 686
+    },
+    {
+      "epoch": 0.1547863801503929,
+      "grad_norm": 0.46846923934248047,
+      "learning_rate": 4.851418435608919e-06,
+      "loss": 0.1802,
+      "step": 687
+    },
+    {
+      "epoch": 0.1550116878362013,
+      "grad_norm": 0.4798823473223289,
+      "learning_rate": 4.850785129069139e-06,
+      "loss": 0.1799,
+      "step": 688
+    },
+    {
+      "epoch": 0.15523699552200976,
+      "grad_norm": 0.47986210011175,
+      "learning_rate": 4.850150517225999e-06,
+      "loss": 0.1797,
+      "step": 689
+    },
+    {
+      "epoch": 0.15546230320781818,
+      "grad_norm": 0.487039403659413,
+      "learning_rate": 4.849514600431877e-06,
+      "loss": 0.1751,
+      "step": 690
+    },
+    {
+      "epoch": 0.15568761089362662,
+      "grad_norm": 0.5225794971495527,
+      "learning_rate": 4.848877379039874e-06,
+      "loss": 0.197,
+      "step": 691
+    },
+    {
+      "epoch": 0.15591291857943504,
+      "grad_norm": 0.48848225992863903,
+      "learning_rate": 4.848238853403813e-06,
+      "loss": 0.1862,
+      "step": 692
+    },
+    {
+      "epoch": 0.15613822626524348,
+      "grad_norm": 0.512252643675161,
+      "learning_rate": 4.847599023878245e-06,
+      "loss": 0.1908,
+      "step": 693
+    },
+    {
+      "epoch": 0.1563635339510519,
+      "grad_norm": 0.47856894671061945,
+      "learning_rate": 4.846957890818444e-06,
+      "loss": 0.1895,
+      "step": 694
+    },
+    {
+      "epoch": 0.15658884163686035,
+      "grad_norm": 0.46385857470598735,
+      "learning_rate": 4.846315454580406e-06,
+      "loss": 0.1796,
+      "step": 695
+    },
+    {
+      "epoch": 0.15681414932266877,
+      "grad_norm": 0.4768661328540527,
+      "learning_rate": 4.845671715520853e-06,
+      "loss": 0.181,
+      "step": 696
+    },
+    {
+      "epoch": 0.1570394570084772,
+      "grad_norm": 0.4603458107546864,
+      "learning_rate": 4.845026673997229e-06,
+      "loss": 0.1729,
+      "step": 697
+    },
+    {
+      "epoch": 0.15726476469428563,
+      "grad_norm": 0.4829231227643716,
+      "learning_rate": 4.844380330367701e-06,
+      "loss": 0.1846,
+      "step": 698
+    },
+    {
+      "epoch": 0.15749007238009408,
+      "grad_norm": 0.5013746953211577,
+      "learning_rate": 4.843732684991161e-06,
+      "loss": 0.1892,
+      "step": 699
+    },
+    {
+      "epoch": 0.1577153800659025,
+      "grad_norm": 0.5322417230269074,
+      "learning_rate": 4.84308373822722e-06,
+      "loss": 0.2024,
+      "step": 700
+    },
+    {
+      "epoch": 0.15794068775171094,
+      "grad_norm": 0.4811274861286996,
+      "learning_rate": 4.842433490436217e-06,
+      "loss": 0.178,
+      "step": 701
+    },
+    {
+      "epoch": 0.15816599543751936,
+      "grad_norm": 0.48964366249033114,
+      "learning_rate": 4.841781941979207e-06,
+      "loss": 0.1952,
+      "step": 702
+    },
+    {
+      "epoch": 0.1583913031233278,
+      "grad_norm": 0.49770700891913777,
+      "learning_rate": 4.8411290932179734e-06,
+      "loss": 0.1836,
+      "step": 703
+    },
+    {
+      "epoch": 0.15861661080913622,
+      "grad_norm": 0.4773017770591324,
+      "learning_rate": 4.840474944515017e-06,
+      "loss": 0.1874,
+      "step": 704
+    },
+    {
+      "epoch": 0.15884191849494467,
+      "grad_norm": 0.519710604598844,
+      "learning_rate": 4.839819496233562e-06,
+      "loss": 0.1914,
+      "step": 705
+    },
+    {
+      "epoch": 0.15906722618075309,
+      "grad_norm": 0.4842922130946162,
+      "learning_rate": 4.839162748737556e-06,
+      "loss": 0.1822,
+      "step": 706
+    },
+    {
+      "epoch": 0.15929253386656153,
+      "grad_norm": 0.5118687486652845,
+      "learning_rate": 4.838504702391665e-06,
+      "loss": 0.1861,
+      "step": 707
+    },
+    {
+      "epoch": 0.15951784155236995,
+      "grad_norm": 0.4944719362975502,
+      "learning_rate": 4.8378453575612785e-06,
+      "loss": 0.1975,
+      "step": 708
+    },
+    {
+      "epoch": 0.1597431492381784,
+      "grad_norm": 0.4645568912744376,
+      "learning_rate": 4.837184714612506e-06,
+      "loss": 0.1745,
+      "step": 709
+    },
+    {
+      "epoch": 0.1599684569239868,
+      "grad_norm": 0.47279405579710293,
+      "learning_rate": 4.836522773912178e-06,
+      "loss": 0.1779,
+      "step": 710
+    },
+    {
+      "epoch": 0.16019376460979526,
+      "grad_norm": 0.5222576281593992,
+      "learning_rate": 4.835859535827844e-06,
+      "loss": 0.1875,
+      "step": 711
+    },
+    {
+      "epoch": 0.16041907229560368,
+      "grad_norm": 0.542641983491453,
+      "learning_rate": 4.835195000727778e-06,
+      "loss": 0.1942,
+      "step": 712
+    },
+    {
+      "epoch": 0.16064437998141212,
+      "grad_norm": 0.5250273856847852,
+      "learning_rate": 4.834529168980969e-06,
+      "loss": 0.1933,
+      "step": 713
+    },
+    {
+      "epoch": 0.16086968766722054,
+      "grad_norm": 0.48171719680672287,
+      "learning_rate": 4.83386204095713e-06,
+      "loss": 0.1853,
+      "step": 714
+    },
+    {
+      "epoch": 0.16109499535302899,
+      "grad_norm": 0.4851873781747077,
+      "learning_rate": 4.833193617026692e-06,
+      "loss": 0.1847,
+      "step": 715
+    },
+    {
+      "epoch": 0.1613203030388374,
+      "grad_norm": 0.5027455855127103,
+      "learning_rate": 4.832523897560806e-06,
+      "loss": 0.1756,
+      "step": 716
+    },
+    {
+      "epoch": 0.16154561072464585,
+      "grad_norm": 0.4754560532712098,
+      "learning_rate": 4.831852882931342e-06,
+      "loss": 0.1824,
+      "step": 717
+    },
+    {
+      "epoch": 0.16177091841045427,
+      "grad_norm": 0.49960393612057585,
+      "learning_rate": 4.83118057351089e-06,
+      "loss": 0.1909,
+      "step": 718
+    },
+    {
+      "epoch": 0.1619962260962627,
+      "grad_norm": 0.4715628869762656,
+      "learning_rate": 4.830506969672758e-06,
+      "loss": 0.1803,
+      "step": 719
+    },
+    {
+      "epoch": 0.16222153378207113,
+      "grad_norm": 0.5079866257652638,
+      "learning_rate": 4.829832071790972e-06,
+      "loss": 0.1946,
+      "step": 720
+    },
+    {
+      "epoch": 0.16244684146787958,
+      "grad_norm": 0.4964010052003955,
+      "learning_rate": 4.829155880240279e-06,
+      "loss": 0.1897,
+      "step": 721
+    },
+    {
+      "epoch": 0.162672149153688,
+      "grad_norm": 0.488861953463059,
+      "learning_rate": 4.828478395396143e-06,
+      "loss": 0.1789,
+      "step": 722
+    },
+    {
+      "epoch": 0.16289745683949644,
+      "grad_norm": 0.47540675122642734,
+      "learning_rate": 4.8277996176347465e-06,
+      "loss": 0.1727,
+      "step": 723
+    },
+    {
+      "epoch": 0.16312276452530486,
+      "grad_norm": 0.4492582480994961,
+      "learning_rate": 4.827119547332988e-06,
+      "loss": 0.1736,
+      "step": 724
+    },
+    {
+      "epoch": 0.1633480722111133,
+      "grad_norm": 0.493647328406061,
+      "learning_rate": 4.826438184868486e-06,
+      "loss": 0.1814,
+      "step": 725
+    },
+    {
+      "epoch": 0.16357337989692172,
+      "grad_norm": 0.4902518098783279,
+      "learning_rate": 4.825755530619576e-06,
+      "loss": 0.1799,
+      "step": 726
+    },
+    {
+      "epoch": 0.16379868758273017,
+      "grad_norm": 0.46936919297661067,
+      "learning_rate": 4.825071584965308e-06,
+      "loss": 0.1826,
+      "step": 727
+    },
+    {
+      "epoch": 0.16402399526853859,
+      "grad_norm": 0.5080991373094106,
+      "learning_rate": 4.824386348285456e-06,
+      "loss": 0.1945,
+      "step": 728
+    },
+    {
+      "epoch": 0.16424930295434703,
+      "grad_norm": 0.5139713054072338,
+      "learning_rate": 4.823699820960502e-06,
+      "loss": 0.1903,
+      "step": 729
+    },
+    {
+      "epoch": 0.16447461064015545,
+      "grad_norm": 0.47537280678853283,
+      "learning_rate": 4.8230120033716525e-06,
+      "loss": 0.1824,
+      "step": 730
+    },
+    {
+      "epoch": 0.1646999183259639,
+      "grad_norm": 0.4648039961115506,
+      "learning_rate": 4.822322895900825e-06,
+      "loss": 0.1739,
+      "step": 731
+    },
+    {
+      "epoch": 0.1649252260117723,
+      "grad_norm": 0.47135193169677986,
+      "learning_rate": 4.821632498930656e-06,
+      "loss": 0.1713,
+      "step": 732
+    },
+    {
+      "epoch": 0.16515053369758076,
+      "grad_norm": 0.4664425826746682,
+      "learning_rate": 4.820940812844496e-06,
+      "loss": 0.1737,
+      "step": 733
+    },
+    {
+      "epoch": 0.16537584138338918,
+      "grad_norm": 0.4884241535946506,
+      "learning_rate": 4.820247838026414e-06,
+      "loss": 0.1822,
+      "step": 734
+    },
+    {
+      "epoch": 0.16560114906919762,
+      "grad_norm": 0.4570363476774046,
+      "learning_rate": 4.819553574861192e-06,
+      "loss": 0.1656,
+      "step": 735
+    },
+    {
+      "epoch": 0.16582645675500607,
+      "grad_norm": 0.5376549865137639,
+      "learning_rate": 4.81885802373433e-06,
+      "loss": 0.1727,
+      "step": 736
+    },
+    {
+      "epoch": 0.1660517644408145,
+      "grad_norm": 0.5108995349206131,
+      "learning_rate": 4.818161185032039e-06,
+      "loss": 0.1998,
+      "step": 737
+    },
+    {
+      "epoch": 0.16627707212662293,
+      "grad_norm": 0.5079851929625643,
+      "learning_rate": 4.8174630591412495e-06,
+      "loss": 0.1982,
+      "step": 738
+    },
+    {
+      "epoch": 0.16650237981243135,
+      "grad_norm": 0.46933454329148294,
+      "learning_rate": 4.816763646449605e-06,
+      "loss": 0.1876,
+      "step": 739
+    },
+    {
+      "epoch": 0.1667276874982398,
+      "grad_norm": 0.47786743858752834,
+      "learning_rate": 4.816062947345462e-06,
+      "loss": 0.1806,
+      "step": 740
+    },
+    {
+      "epoch": 0.16695299518404821,
+      "grad_norm": 0.5149184082938212,
+      "learning_rate": 4.815360962217894e-06,
+      "loss": 0.2055,
+      "step": 741
+    },
+    {
+      "epoch": 0.16717830286985666,
+      "grad_norm": 0.4839519889099994,
+      "learning_rate": 4.814657691456685e-06,
+      "loss": 0.1807,
+      "step": 742
+    },
+    {
+      "epoch": 0.16740361055566508,
+      "grad_norm": 0.5097448035834458,
+      "learning_rate": 4.813953135452338e-06,
+      "loss": 0.1883,
+      "step": 743
+    },
+    {
+      "epoch": 0.16762891824147352,
+      "grad_norm": 0.49981444304743583,
+      "learning_rate": 4.813247294596065e-06,
+      "loss": 0.1873,
+      "step": 744
+    },
+    {
+      "epoch": 0.16785422592728194,
+      "grad_norm": 0.49909466388052154,
+      "learning_rate": 4.812540169279793e-06,
+      "loss": 0.184,
+      "step": 745
+    },
+    {
+      "epoch": 0.1680795336130904,
+      "grad_norm": 0.4853088099017004,
+      "learning_rate": 4.8118317598961625e-06,
+      "loss": 0.1828,
+      "step": 746
+    },
+    {
+      "epoch": 0.1683048412988988,
+      "grad_norm": 0.512848849396602,
+      "learning_rate": 4.811122066838527e-06,
+      "loss": 0.1743,
+      "step": 747
+    },
+    {
+      "epoch": 0.16853014898470725,
+      "grad_norm": 0.4945737090434059,
+      "learning_rate": 4.810411090500952e-06,
+      "loss": 0.1841,
+      "step": 748
+    },
+    {
+      "epoch": 0.16875545667051567,
+      "grad_norm": 0.484533971919501,
+      "learning_rate": 4.809698831278217e-06,
+      "loss": 0.1681,
+      "step": 749
+    },
+    {
+      "epoch": 0.16898076435632411,
+      "grad_norm": 0.5274434401849899,
+      "learning_rate": 4.808985289565813e-06,
+      "loss": 0.2138,
+      "step": 750
+    },
+    {
+      "epoch": 0.16920607204213253,
+      "grad_norm": 0.5429817492746462,
+      "learning_rate": 4.808270465759943e-06,
+      "loss": 0.1932,
+      "step": 751
+    },
+    {
+      "epoch": 0.16943137972794098,
+      "grad_norm": 0.4898706441354639,
+      "learning_rate": 4.807554360257522e-06,
+      "loss": 0.2004,
+      "step": 752
+    },
+    {
+      "epoch": 0.1696566874137494,
+      "grad_norm": 0.49637968065289056,
+      "learning_rate": 4.806836973456175e-06,
+      "loss": 0.1822,
+      "step": 753
+    },
+    {
+      "epoch": 0.16988199509955784,
+      "grad_norm": 0.5084117755419836,
+      "learning_rate": 4.8061183057542424e-06,
+      "loss": 0.1715,
+      "step": 754
+    },
+    {
+      "epoch": 0.17010730278536626,
+      "grad_norm": 0.5109495087867058,
+      "learning_rate": 4.8053983575507735e-06,
+      "loss": 0.1765,
+      "step": 755
+    },
+    {
+      "epoch": 0.1703326104711747,
+      "grad_norm": 0.4928451254071443,
+      "learning_rate": 4.804677129245527e-06,
+      "loss": 0.1859,
+      "step": 756
+    },
+    {
+      "epoch": 0.17055791815698312,
+      "grad_norm": 0.46419293537472645,
+      "learning_rate": 4.8039546212389765e-06,
+      "loss": 0.1706,
+      "step": 757
+    },
+    {
+      "epoch": 0.17078322584279157,
+      "grad_norm": 0.4830003007409314,
+      "learning_rate": 4.803230833932302e-06,
+      "loss": 0.1832,
+      "step": 758
+    },
+    {
+      "epoch": 0.1710085335286,
+      "grad_norm": 0.46911559787991525,
+      "learning_rate": 4.802505767727395e-06,
+      "loss": 0.1775,
+      "step": 759
+    },
+    {
+      "epoch": 0.17123384121440843,
+      "grad_norm": 0.5119510412459805,
+      "learning_rate": 4.80177942302686e-06,
+      "loss": 0.1921,
+      "step": 760
+    },
+    {
+      "epoch": 0.17145914890021685,
+      "grad_norm": 0.4866339023029943,
+      "learning_rate": 4.8010518002340065e-06,
+      "loss": 0.1753,
+      "step": 761
+    },
+    {
+      "epoch": 0.1716844565860253,
+      "grad_norm": 0.4615836386800967,
+      "learning_rate": 4.800322899752859e-06,
+      "loss": 0.1692,
+      "step": 762
+    },
+    {
+      "epoch": 0.17190976427183371,
+      "grad_norm": 0.5165845262276799,
+      "learning_rate": 4.799592721988147e-06,
+      "loss": 0.1963,
+      "step": 763
+    },
+    {
+      "epoch": 0.17213507195764216,
+      "grad_norm": 0.45977940689859065,
+      "learning_rate": 4.798861267345312e-06,
+      "loss": 0.1699,
+      "step": 764
+    },
+    {
+      "epoch": 0.17236037964345058,
+      "grad_norm": 0.4787043266174051,
+      "learning_rate": 4.798128536230502e-06,
+      "loss": 0.1871,
+      "step": 765
+    },
+    {
+      "epoch": 0.17258568732925902,
+      "grad_norm": 0.4888347653144028,
+      "learning_rate": 4.797394529050577e-06,
+      "loss": 0.1896,
+      "step": 766
+    },
+    {
+      "epoch": 0.17281099501506744,
+      "grad_norm": 0.4914002260078396,
+      "learning_rate": 4.796659246213103e-06,
+      "loss": 0.1819,
+      "step": 767
+    },
+    {
+      "epoch": 0.1730363027008759,
+      "grad_norm": 0.4731805055670691,
+      "learning_rate": 4.795922688126355e-06,
+      "loss": 0.1655,
+      "step": 768
+    },
+    {
+      "epoch": 0.1732616103866843,
+      "grad_norm": 0.5111009862183389,
+      "learning_rate": 4.795184855199316e-06,
+      "loss": 0.2014,
+      "step": 769
+    },
+    {
+      "epoch": 0.17348691807249275,
+      "grad_norm": 0.5137505817476531,
+      "learning_rate": 4.794445747841679e-06,
+      "loss": 0.1966,
+      "step": 770
+    },
+    {
+      "epoch": 0.17371222575830117,
+      "grad_norm": 0.50296528263997,
+      "learning_rate": 4.79370536646384e-06,
+      "loss": 0.1944,
+      "step": 771
+    },
+    {
+      "epoch": 0.17393753344410962,
+      "grad_norm": 0.48127568016573596,
+      "learning_rate": 4.792963711476908e-06,
+      "loss": 0.1809,
+      "step": 772
+    },
+    {
+      "epoch": 0.17416284112991803,
+      "grad_norm": 0.4572815368441324,
+      "learning_rate": 4.792220783292694e-06,
+      "loss": 0.1683,
+      "step": 773
+    },
+    {
+      "epoch": 0.17438814881572648,
+      "grad_norm": 0.48725892749917493,
+      "learning_rate": 4.791476582323719e-06,
+      "loss": 0.179,
+      "step": 774
+    },
+    {
+      "epoch": 0.1746134565015349,
+      "grad_norm": 0.49300456728333963,
+      "learning_rate": 4.790731108983211e-06,
+      "loss": 0.184,
+      "step": 775
+    },
+    {
+      "epoch": 0.17483876418734334,
+      "grad_norm": 0.4787596622664503,
+      "learning_rate": 4.7899843636851014e-06,
+      "loss": 0.1717,
+      "step": 776
+    },
+    {
+      "epoch": 0.17506407187315176,
+      "grad_norm": 0.48381195789429615,
+      "learning_rate": 4.789236346844034e-06,
+      "loss": 0.1835,
+      "step": 777
+    },
+    {
+      "epoch": 0.1752893795589602,
+      "grad_norm": 0.4933319234608218,
+      "learning_rate": 4.78848705887535e-06,
+      "loss": 0.184,
+      "step": 778
+    },
+    {
+      "epoch": 0.17551468724476862,
+      "grad_norm": 0.5095035300039361,
+      "learning_rate": 4.7877365001951045e-06,
+      "loss": 0.1809,
+      "step": 779
+    },
+    {
+      "epoch": 0.17573999493057707,
+      "grad_norm": 0.4940075922499647,
+      "learning_rate": 4.786984671220053e-06,
+      "loss": 0.1647,
+      "step": 780
+    },
+    {
+      "epoch": 0.1759653026163855,
+      "grad_norm": 0.5158596908756363,
+      "learning_rate": 4.786231572367659e-06,
+      "loss": 0.1945,
+      "step": 781
+    },
+    {
+      "epoch": 0.17619061030219393,
+      "grad_norm": 0.462830581703408,
+      "learning_rate": 4.785477204056089e-06,
+      "loss": 0.1778,
+      "step": 782
+    },
+    {
+      "epoch": 0.17641591798800235,
+      "grad_norm": 0.5339185435557559,
+      "learning_rate": 4.784721566704217e-06,
+      "loss": 0.187,
+      "step": 783
+    },
+    {
+      "epoch": 0.1766412256738108,
+      "grad_norm": 0.5257175194638931,
+      "learning_rate": 4.78396466073162e-06,
+      "loss": 0.1816,
+      "step": 784
+    },
+    {
+      "epoch": 0.17686653335961924,
+      "grad_norm": 0.5028416895616926,
+      "learning_rate": 4.7832064865585795e-06,
+      "loss": 0.1794,
+      "step": 785
+    },
+    {
+      "epoch": 0.17709184104542766,
+      "grad_norm": 0.47469854396033484,
+      "learning_rate": 4.78244704460608e-06,
+      "loss": 0.1762,
+      "step": 786
+    },
+    {
+      "epoch": 0.1773171487312361,
+      "grad_norm": 0.5164293026713752,
+      "learning_rate": 4.781686335295813e-06,
+      "loss": 0.1823,
+      "step": 787
+    },
+    {
+      "epoch": 0.17754245641704453,
+      "grad_norm": 0.5341407689751599,
+      "learning_rate": 4.7809243590501725e-06,
+      "loss": 0.1819,
+      "step": 788
+    },
+    {
+      "epoch": 0.17776776410285297,
+      "grad_norm": 0.5041897167760184,
+      "learning_rate": 4.780161116292254e-06,
+      "loss": 0.1838,
+      "step": 789
+    },
+    {
+      "epoch": 0.1779930717886614,
+      "grad_norm": 0.4777200934140291,
+      "learning_rate": 4.779396607445858e-06,
+      "loss": 0.1859,
+      "step": 790
+    },
+    {
+      "epoch": 0.17821837947446983,
+      "grad_norm": 0.5105523444886348,
+      "learning_rate": 4.778630832935489e-06,
+      "loss": 0.18,
+      "step": 791
+    },
+    {
+      "epoch": 0.17844368716027825,
+      "grad_norm": 0.5049092473527552,
+      "learning_rate": 4.777863793186351e-06,
+      "loss": 0.1772,
+      "step": 792
+    },
+    {
+      "epoch": 0.1786689948460867,
+      "grad_norm": 0.47907869755384375,
+      "learning_rate": 4.777095488624355e-06,
+      "loss": 0.1684,
+      "step": 793
+    },
+    {
+      "epoch": 0.17889430253189512,
+      "grad_norm": 0.5311995304246812,
+      "learning_rate": 4.776325919676109e-06,
+      "loss": 0.1981,
+      "step": 794
+    },
+    {
+      "epoch": 0.17911961021770356,
+      "grad_norm": 0.48060992323771873,
+      "learning_rate": 4.775555086768929e-06,
+      "loss": 0.1708,
+      "step": 795
+    },
+    {
+      "epoch": 0.17934491790351198,
+      "grad_norm": 0.4675293098501905,
+      "learning_rate": 4.774782990330828e-06,
+      "loss": 0.1703,
+      "step": 796
+    },
+    {
+      "epoch": 0.17957022558932043,
+      "grad_norm": 0.4930964788341703,
+      "learning_rate": 4.774009630790522e-06,
+      "loss": 0.1921,
+      "step": 797
+    },
+    {
+      "epoch": 0.17979553327512884,
+      "grad_norm": 0.46495600709912344,
+      "learning_rate": 4.77323500857743e-06,
+      "loss": 0.1599,
+      "step": 798
+    },
+    {
+      "epoch": 0.1800208409609373,
+      "grad_norm": 0.47676582472244317,
+      "learning_rate": 4.77245912412167e-06,
+      "loss": 0.1834,
+      "step": 799
+    },
+    {
+      "epoch": 0.1802461486467457,
+      "grad_norm": 0.4895979712857663,
+      "learning_rate": 4.771681977854062e-06,
+      "loss": 0.1761,
+      "step": 800
+    },
+    {
+      "epoch": 0.18047145633255415,
+      "grad_norm": 0.4932387130749087,
+      "learning_rate": 4.7709035702061275e-06,
+      "loss": 0.1781,
+      "step": 801
+    },
+    {
+      "epoch": 0.18069676401836257,
+      "grad_norm": 0.4779913850319623,
+      "learning_rate": 4.770123901610085e-06,
+      "loss": 0.1824,
+      "step": 802
+    },
+    {
+      "epoch": 0.18092207170417102,
+      "grad_norm": 0.48771377451063314,
+      "learning_rate": 4.7693429724988565e-06,
+      "loss": 0.1819,
+      "step": 803
+    },
+    {
+      "epoch": 0.18114737938997943,
+      "grad_norm": 0.48970848842547954,
+      "learning_rate": 4.768560783306064e-06,
+      "loss": 0.1903,
+      "step": 804
+    },
+    {
+      "epoch": 0.18137268707578788,
+      "grad_norm": 0.45068406552687224,
+      "learning_rate": 4.767777334466025e-06,
+      "loss": 0.1704,
+      "step": 805
+    },
+    {
+      "epoch": 0.1815979947615963,
+      "grad_norm": 0.5051420899177393,
+      "learning_rate": 4.7669926264137625e-06,
+      "loss": 0.1837,
+      "step": 806
+    },
+    {
+      "epoch": 0.18182330244740474,
+      "grad_norm": 0.48811515671301603,
+      "learning_rate": 4.766206659584994e-06,
+      "loss": 0.1641,
+      "step": 807
+    },
+    {
+      "epoch": 0.18204861013321316,
+      "grad_norm": 0.4851157603919634,
+      "learning_rate": 4.765419434416138e-06,
+      "loss": 0.1825,
+      "step": 808
+    },
+    {
+      "epoch": 0.1822739178190216,
+      "grad_norm": 0.5082718059063132,
+      "learning_rate": 4.7646309513443115e-06,
+      "loss": 0.1985,
+      "step": 809
+    },
+    {
+      "epoch": 0.18249922550483003,
+      "grad_norm": 0.4791172574916238,
+      "learning_rate": 4.763841210807329e-06,
+      "loss": 0.1722,
+      "step": 810
+    },
+    {
+      "epoch": 0.18272453319063847,
+      "grad_norm": 0.4901673009180037,
+      "learning_rate": 4.763050213243705e-06,
+      "loss": 0.1835,
+      "step": 811
+    },
+    {
+      "epoch": 0.1829498408764469,
+      "grad_norm": 0.49925031809883547,
+      "learning_rate": 4.762257959092651e-06,
+      "loss": 0.1934,
+      "step": 812
+    },
+    {
+      "epoch": 0.18317514856225534,
+      "grad_norm": 0.46501080227779795,
+      "learning_rate": 4.7614644487940755e-06,
+      "loss": 0.1817,
+      "step": 813
+    },
+    {
+      "epoch": 0.18340045624806375,
+      "grad_norm": 0.483217674981779,
+      "learning_rate": 4.760669682788584e-06,
+      "loss": 0.1872,
+      "step": 814
+    },
+    {
+      "epoch": 0.1836257639338722,
+      "grad_norm": 0.5447110598554856,
+      "learning_rate": 4.759873661517484e-06,
+      "loss": 0.1946,
+      "step": 815
+    },
+    {
+      "epoch": 0.18385107161968062,
+      "grad_norm": 0.4963607265926464,
+      "learning_rate": 4.759076385422773e-06,
+      "loss": 0.1673,
+      "step": 816
+    },
+    {
+      "epoch": 0.18407637930548906,
+      "grad_norm": 0.49715295354011035,
+      "learning_rate": 4.75827785494715e-06,
+      "loss": 0.1822,
+      "step": 817
+    },
+    {
+      "epoch": 0.18430168699129748,
+      "grad_norm": 0.4468255135593684,
+      "learning_rate": 4.7574780705340094e-06,
+      "loss": 0.1664,
+      "step": 818
+    },
+    {
+      "epoch": 0.18452699467710593,
+      "grad_norm": 0.5020265233870463,
+      "learning_rate": 4.756677032627442e-06,
+      "loss": 0.189,
+      "step": 819
+    },
+    {
+      "epoch": 0.18475230236291434,
+      "grad_norm": 0.49870372725644924,
+      "learning_rate": 4.755874741672233e-06,
+      "loss": 0.1778,
+      "step": 820
+    },
+    {
+      "epoch": 0.1849776100487228,
+      "grad_norm": 0.457171013269724,
+      "learning_rate": 4.755071198113865e-06,
+      "loss": 0.1725,
+      "step": 821
+    },
+    {
+      "epoch": 0.1852029177345312,
+      "grad_norm": 0.5092879280720725,
+      "learning_rate": 4.754266402398517e-06,
+      "loss": 0.1813,
+      "step": 822
+    },
+    {
+      "epoch": 0.18542822542033965,
+      "grad_norm": 0.5166784139586021,
+      "learning_rate": 4.753460354973061e-06,
+      "loss": 0.1866,
+      "step": 823
+    },
+    {
+      "epoch": 0.18565353310614807,
+      "grad_norm": 0.4861158761653109,
+      "learning_rate": 4.752653056285066e-06,
+      "loss": 0.1717,
+      "step": 824
+    },
+    {
+      "epoch": 0.18587884079195652,
+      "grad_norm": 0.46905214327282685,
+      "learning_rate": 4.751844506782793e-06,
+      "loss": 0.1862,
+      "step": 825
+    },
+    {
+      "epoch": 0.18610414847776494,
+      "grad_norm": 0.44862810839169803,
+      "learning_rate": 4.7510347069152015e-06,
+      "loss": 0.1644,
+      "step": 826
+    },
+    {
+      "epoch": 0.18632945616357338,
+      "grad_norm": 0.4975474972517688,
+      "learning_rate": 4.750223657131942e-06,
+      "loss": 0.1822,
+      "step": 827
+    },
+    {
+      "epoch": 0.1865547638493818,
+      "grad_norm": 0.4765117099823277,
+      "learning_rate": 4.74941135788336e-06,
+      "loss": 0.1779,
+      "step": 828
+    },
+    {
+      "epoch": 0.18678007153519025,
+      "grad_norm": 0.45414080502387294,
+      "learning_rate": 4.748597809620496e-06,
+      "loss": 0.1569,
+      "step": 829
+    },
+    {
+      "epoch": 0.18700537922099866,
+      "grad_norm": 0.5041355859519643,
+      "learning_rate": 4.747783012795083e-06,
+      "loss": 0.1986,
+      "step": 830
+    },
+    {
+      "epoch": 0.1872306869068071,
+      "grad_norm": 0.5051258971770674,
+      "learning_rate": 4.746966967859547e-06,
+      "loss": 0.1951,
+      "step": 831
+    },
+    {
+      "epoch": 0.18745599459261553,
+      "grad_norm": 0.49081074092309795,
+      "learning_rate": 4.746149675267005e-06,
+      "loss": 0.1804,
+      "step": 832
+    },
+    {
+      "epoch": 0.18768130227842397,
+      "grad_norm": 0.4441868113966922,
+      "learning_rate": 4.745331135471274e-06,
+      "loss": 0.1585,
+      "step": 833
+    },
+    {
+      "epoch": 0.18790660996423242,
+      "grad_norm": 0.46703122083044185,
+      "learning_rate": 4.744511348926855e-06,
+      "loss": 0.176,
+      "step": 834
+    },
+    {
+      "epoch": 0.18813191765004084,
+      "grad_norm": 0.5208426656798611,
+      "learning_rate": 4.743690316088945e-06,
+      "loss": 0.1815,
+      "step": 835
+    },
+    {
+      "epoch": 0.18835722533584928,
+      "grad_norm": 0.5164756735393626,
+      "learning_rate": 4.742868037413435e-06,
+      "loss": 0.1972,
+      "step": 836
+    },
+    {
+      "epoch": 0.1885825330216577,
+      "grad_norm": 0.49649766308699994,
+      "learning_rate": 4.742044513356904e-06,
+      "loss": 0.1754,
+      "step": 837
+    },
+    {
+      "epoch": 0.18880784070746615,
+      "grad_norm": 0.49273564586087437,
+      "learning_rate": 4.741219744376624e-06,
+      "loss": 0.1712,
+      "step": 838
+    },
+    {
+      "epoch": 0.18903314839327456,
+      "grad_norm": 0.5486282244513546,
+      "learning_rate": 4.74039373093056e-06,
+      "loss": 0.2004,
+      "step": 839
+    },
+    {
+      "epoch": 0.189258456079083,
+      "grad_norm": 0.4748075709655306,
+      "learning_rate": 4.739566473477365e-06,
+      "loss": 0.1686,
+      "step": 840
+    },
+    {
+      "epoch": 0.18948376376489143,
+      "grad_norm": 0.48522772612578585,
+      "learning_rate": 4.738737972476385e-06,
+      "loss": 0.1665,
+      "step": 841
+    },
+    {
+      "epoch": 0.18970907145069987,
+      "grad_norm": 0.5329091938352456,
+      "learning_rate": 4.737908228387656e-06,
+      "loss": 0.1864,
+      "step": 842
+    },
+    {
+      "epoch": 0.1899343791365083,
+      "grad_norm": 0.5047847276742794,
+      "learning_rate": 4.737077241671904e-06,
+      "loss": 0.1929,
+      "step": 843
+    },
+    {
+      "epoch": 0.19015968682231674,
+      "grad_norm": 0.46222373907325087,
+      "learning_rate": 4.736245012790543e-06,
+      "loss": 0.1741,
+      "step": 844
+    },
+    {
+      "epoch": 0.19038499450812515,
+      "grad_norm": 0.5185005587102809,
+      "learning_rate": 4.735411542205681e-06,
+      "loss": 0.1912,
+      "step": 845
+    },
+    {
+      "epoch": 0.1906103021939336,
+      "grad_norm": 0.5116459064135198,
+      "learning_rate": 4.734576830380113e-06,
+      "loss": 0.176,
+      "step": 846
+    },
+    {
+      "epoch": 0.19083560987974202,
+      "grad_norm": 0.4666206667887288,
+      "learning_rate": 4.733740877777322e-06,
+      "loss": 0.1706,
+      "step": 847
+    },
+    {
+      "epoch": 0.19106091756555046,
+      "grad_norm": 0.4967074435816717,
+      "learning_rate": 4.732903684861482e-06,
+      "loss": 0.1873,
+      "step": 848
+    },
+    {
+      "epoch": 0.19128622525135888,
+      "grad_norm": 0.5365788421064641,
+      "learning_rate": 4.732065252097455e-06,
+      "loss": 0.1847,
+      "step": 849
+    },
+    {
+      "epoch": 0.19151153293716733,
+      "grad_norm": 0.4977595692595492,
+      "learning_rate": 4.731225579950791e-06,
+      "loss": 0.1698,
+      "step": 850
+    },
+    {
+      "epoch": 0.19173684062297575,
+      "grad_norm": 0.5220653425157609,
+      "learning_rate": 4.730384668887731e-06,
+      "loss": 0.1858,
+      "step": 851
+    },
+    {
+      "epoch": 0.1919621483087842,
+      "grad_norm": 0.48911620839265174,
+      "learning_rate": 4.7295425193751974e-06,
+      "loss": 0.1704,
+      "step": 852
+    },
+    {
+      "epoch": 0.1921874559945926,
+      "grad_norm": 0.5340840544161292,
+      "learning_rate": 4.728699131880808e-06,
+      "loss": 0.1962,
+      "step": 853
+    },
+    {
+      "epoch": 0.19241276368040106,
+      "grad_norm": 0.4885150418015092,
+      "learning_rate": 4.727854506872863e-06,
+      "loss": 0.1618,
+      "step": 854
+    },
+    {
+      "epoch": 0.19263807136620947,
+      "grad_norm": 0.549170399664639,
+      "learning_rate": 4.727008644820351e-06,
+      "loss": 0.1914,
+      "step": 855
+    },
+    {
+      "epoch": 0.19286337905201792,
+      "grad_norm": 0.4526719253769709,
+      "learning_rate": 4.726161546192949e-06,
+      "loss": 0.169,
+      "step": 856
+    },
+    {
+      "epoch": 0.19308868673782634,
+      "grad_norm": 0.5149778393372377,
+      "learning_rate": 4.725313211461018e-06,
+      "loss": 0.1922,
+      "step": 857
+    },
+    {
+      "epoch": 0.19331399442363478,
+      "grad_norm": 0.5379703316294756,
+      "learning_rate": 4.724463641095606e-06,
+      "loss": 0.1789,
+      "step": 858
+    },
+    {
+      "epoch": 0.1935393021094432,
+      "grad_norm": 0.4923803644073318,
+      "learning_rate": 4.72361283556845e-06,
+      "loss": 0.1811,
+      "step": 859
+    },
+    {
+      "epoch": 0.19376460979525165,
+      "grad_norm": 0.48133842551153155,
+      "learning_rate": 4.7227607953519686e-06,
+      "loss": 0.1743,
+      "step": 860
+    },
+    {
+      "epoch": 0.19398991748106006,
+      "grad_norm": 0.5108863471576068,
+      "learning_rate": 4.7219075209192686e-06,
+      "loss": 0.1844,
+      "step": 861
+    },
+    {
+      "epoch": 0.1942152251668685,
+      "grad_norm": 0.472380424434701,
+      "learning_rate": 4.721053012744142e-06,
+      "loss": 0.169,
+      "step": 862
+    },
+    {
+      "epoch": 0.19444053285267693,
+      "grad_norm": 0.4919449085715017,
+      "learning_rate": 4.720197271301064e-06,
+      "loss": 0.1816,
+      "step": 863
+    },
+    {
+      "epoch": 0.19466584053848537,
+      "grad_norm": 0.4749560870382914,
+      "learning_rate": 4.719340297065198e-06,
+      "loss": 0.1737,
+      "step": 864
+    },
+    {
+      "epoch": 0.1948911482242938,
+      "grad_norm": 0.5267170058766248,
+      "learning_rate": 4.718482090512389e-06,
+      "loss": 0.187,
+      "step": 865
+    },
+    {
+      "epoch": 0.19511645591010224,
+      "grad_norm": 0.4582779313083357,
+      "learning_rate": 4.717622652119166e-06,
+      "loss": 0.1596,
+      "step": 866
+    },
+    {
+      "epoch": 0.19534176359591066,
+      "grad_norm": 0.4770738218995198,
+      "learning_rate": 4.716761982362744e-06,
+      "loss": 0.1872,
+      "step": 867
+    },
+    {
+      "epoch": 0.1955670712817191,
+      "grad_norm": 0.4935886966800019,
+      "learning_rate": 4.715900081721021e-06,
+      "loss": 0.1841,
+      "step": 868
+    },
+    {
+      "epoch": 0.19579237896752752,
+      "grad_norm": 0.5217854576232532,
+      "learning_rate": 4.715036950672578e-06,
+      "loss": 0.175,
+      "step": 869
+    },
+    {
+      "epoch": 0.19601768665333597,
+      "grad_norm": 0.4785621830298838,
+      "learning_rate": 4.71417258969668e-06,
+      "loss": 0.1786,
+      "step": 870
+    },
+    {
+      "epoch": 0.19624299433914438,
+      "grad_norm": 0.46485530569258604,
+      "learning_rate": 4.713306999273273e-06,
+      "loss": 0.1761,
+      "step": 871
+    },
+    {
+      "epoch": 0.19646830202495283,
+      "grad_norm": 0.44349726343989515,
+      "learning_rate": 4.712440179882989e-06,
+      "loss": 0.1579,
+      "step": 872
+    },
+    {
+      "epoch": 0.19669360971076125,
+      "grad_norm": 0.5228158669046902,
+      "learning_rate": 4.711572132007139e-06,
+      "loss": 0.184,
+      "step": 873
+    },
+    {
+      "epoch": 0.1969189173965697,
+      "grad_norm": 0.4669706169996906,
+      "learning_rate": 4.710702856127718e-06,
+      "loss": 0.1706,
+      "step": 874
+    },
+    {
+      "epoch": 0.1971442250823781,
+      "grad_norm": 0.47828996129737333,
+      "learning_rate": 4.709832352727404e-06,
+      "loss": 0.1643,
+      "step": 875
+    },
+    {
+      "epoch": 0.19736953276818656,
+      "grad_norm": 0.49478309332034437,
+      "learning_rate": 4.708960622289552e-06,
+      "loss": 0.1975,
+      "step": 876
+    },
+    {
+      "epoch": 0.19759484045399497,
+      "grad_norm": 0.5005389974519691,
+      "learning_rate": 4.708087665298204e-06,
+      "loss": 0.1914,
+      "step": 877
+    },
+    {
+      "epoch": 0.19782014813980342,
+      "grad_norm": 0.4596868416283332,
+      "learning_rate": 4.70721348223808e-06,
+      "loss": 0.1817,
+      "step": 878
+    },
+    {
+      "epoch": 0.19804545582561184,
+      "grad_norm": 0.45406849764461193,
+      "learning_rate": 4.706338073594581e-06,
+      "loss": 0.171,
+      "step": 879
+    },
+    {
+      "epoch": 0.19827076351142028,
+      "grad_norm": 0.4994377731966007,
+      "learning_rate": 4.705461439853789e-06,
+      "loss": 0.1893,
+      "step": 880
+    },
+    {
+      "epoch": 0.1984960711972287,
+      "grad_norm": 0.471170874746787,
+      "learning_rate": 4.704583581502465e-06,
+      "loss": 0.1765,
+      "step": 881
+    },
+    {
+      "epoch": 0.19872137888303715,
+      "grad_norm": 0.5304815833710694,
+      "learning_rate": 4.703704499028052e-06,
+      "loss": 0.2032,
+      "step": 882
+    },
+    {
+      "epoch": 0.1989466865688456,
+      "grad_norm": 0.47495896919317876,
+      "learning_rate": 4.702824192918672e-06,
+      "loss": 0.1711,
+      "step": 883
+    },
+    {
+      "epoch": 0.199171994254654,
+      "grad_norm": 0.48490606250503726,
+      "learning_rate": 4.701942663663126e-06,
+      "loss": 0.1733,
+      "step": 884
+    },
+    {
+      "epoch": 0.19939730194046246,
+      "grad_norm": 0.48968905810940944,
+      "learning_rate": 4.7010599117508936e-06,
+      "loss": 0.1752,
+      "step": 885
+    },
+    {
+      "epoch": 0.19962260962627087,
+      "grad_norm": 0.47967576032175524,
+      "learning_rate": 4.700175937672134e-06,
+      "loss": 0.1873,
+      "step": 886
+    },
+    {
+      "epoch": 0.19984791731207932,
+      "grad_norm": 0.5026032086154385,
+      "learning_rate": 4.699290741917686e-06,
+      "loss": 0.173,
+      "step": 887
+    },
+    {
+      "epoch": 0.20007322499788774,
+      "grad_norm": 0.4879917529677437,
+      "learning_rate": 4.698404324979066e-06,
+      "loss": 0.1778,
+      "step": 888
+    },
+    {
+      "epoch": 0.20029853268369618,
+      "grad_norm": 0.476247466553313,
+      "learning_rate": 4.697516687348466e-06,
+      "loss": 0.1812,
+      "step": 889
+    },
+    {
+      "epoch": 0.2005238403695046,
+      "grad_norm": 0.5052080033610629,
+      "learning_rate": 4.696627829518761e-06,
+      "loss": 0.1901,
+      "step": 890
+    },
+    {
+      "epoch": 0.20074914805531305,
+      "grad_norm": 0.48885147290925035,
+      "learning_rate": 4.695737751983499e-06,
+      "loss": 0.1822,
+      "step": 891
+    },
+    {
+      "epoch": 0.20097445574112147,
+      "grad_norm": 0.49482722093377984,
+      "learning_rate": 4.6948464552369075e-06,
+      "loss": 0.1757,
+      "step": 892
+    },
+    {
+      "epoch": 0.2011997634269299,
+      "grad_norm": 0.4734093310501187,
+      "learning_rate": 4.69395393977389e-06,
+      "loss": 0.1755,
+      "step": 893
+    },
+    {
+      "epoch": 0.20142507111273833,
+      "grad_norm": 0.4708301078264215,
+      "learning_rate": 4.693060206090028e-06,
+      "loss": 0.1754,
+      "step": 894
+    },
+    {
+      "epoch": 0.20165037879854678,
+      "grad_norm": 0.4783063902662343,
+      "learning_rate": 4.692165254681576e-06,
+      "loss": 0.1833,
+      "step": 895
+    },
+    {
+      "epoch": 0.2018756864843552,
+      "grad_norm": 0.4884538782786743,
+      "learning_rate": 4.69126908604547e-06,
+      "loss": 0.1793,
+      "step": 896
+    },
+    {
+      "epoch": 0.20210099417016364,
+      "grad_norm": 0.5113830943043448,
+      "learning_rate": 4.690371700679317e-06,
+      "loss": 0.1871,
+      "step": 897
+    },
+    {
+      "epoch": 0.20232630185597206,
+      "grad_norm": 0.48752649520933233,
+      "learning_rate": 4.689473099081403e-06,
+      "loss": 0.1882,
+      "step": 898
+    },
+    {
+      "epoch": 0.2025516095417805,
+      "grad_norm": 0.47456621778022884,
+      "learning_rate": 4.688573281750688e-06,
+      "loss": 0.1612,
+      "step": 899
+    },
+    {
+      "epoch": 0.20277691722758892,
+      "grad_norm": 0.5307489124130947,
+      "learning_rate": 4.687672249186805e-06,
+      "loss": 0.1729,
+      "step": 900
+    },
+    {
+      "epoch": 0.20300222491339737,
+      "grad_norm": 0.4664993822416109,
+      "learning_rate": 4.686770001890067e-06,
+      "loss": 0.163,
+      "step": 901
+    },
+    {
+      "epoch": 0.20322753259920578,
+      "grad_norm": 0.5132544106040071,
+      "learning_rate": 4.685866540361456e-06,
+      "loss": 0.1893,
+      "step": 902
+    },
+    {
+      "epoch": 0.20345284028501423,
+      "grad_norm": 0.5141748664372958,
+      "learning_rate": 4.684961865102631e-06,
+      "loss": 0.1643,
+      "step": 903
+    },
+    {
+      "epoch": 0.20367814797082265,
+      "grad_norm": 0.4955830704478595,
+      "learning_rate": 4.684055976615924e-06,
+      "loss": 0.1639,
+      "step": 904
+    },
+    {
+      "epoch": 0.2039034556566311,
+      "grad_norm": 0.48236349752009133,
+      "learning_rate": 4.683148875404343e-06,
+      "loss": 0.1797,
+      "step": 905
+    },
+    {
+      "epoch": 0.2041287633424395,
+      "grad_norm": 0.5427867503596624,
+      "learning_rate": 4.682240561971565e-06,
+      "loss": 0.1779,
+      "step": 906
+    },
+    {
+      "epoch": 0.20435407102824796,
+      "grad_norm": 0.5251684159747709,
+      "learning_rate": 4.681331036821945e-06,
+      "loss": 0.1873,
+      "step": 907
+    },
+    {
+      "epoch": 0.20457937871405638,
+      "grad_norm": 0.5375306958337258,
+      "learning_rate": 4.680420300460505e-06,
+      "loss": 0.1949,
+      "step": 908
+    },
+    {
+      "epoch": 0.20480468639986482,
+      "grad_norm": 0.46761640654985387,
+      "learning_rate": 4.679508353392946e-06,
+      "loss": 0.1653,
+      "step": 909
+    },
+    {
+      "epoch": 0.20502999408567324,
+      "grad_norm": 0.46505944697352847,
+      "learning_rate": 4.678595196125638e-06,
+      "loss": 0.166,
+      "step": 910
+    },
+    {
+      "epoch": 0.20525530177148169,
+      "grad_norm": 0.47542350967948827,
+      "learning_rate": 4.677680829165623e-06,
+      "loss": 0.1734,
+      "step": 911
+    },
+    {
+      "epoch": 0.2054806094572901,
+      "grad_norm": 0.5038195396548182,
+      "learning_rate": 4.676765253020613e-06,
+      "loss": 0.1685,
+      "step": 912
+    },
+    {
+      "epoch": 0.20570591714309855,
+      "grad_norm": 0.5456800961270842,
+      "learning_rate": 4.675848468198995e-06,
+      "loss": 0.1952,
+      "step": 913
+    },
+    {
+      "epoch": 0.20593122482890697,
+      "grad_norm": 0.47717326012790684,
+      "learning_rate": 4.674930475209827e-06,
+      "loss": 0.1717,
+      "step": 914
+    },
+    {
+      "epoch": 0.2061565325147154,
+      "grad_norm": 0.48833151208415465,
+      "learning_rate": 4.674011274562833e-06,
+      "loss": 0.1624,
+      "step": 915
+    },
+    {
+      "epoch": 0.20638184020052383,
+      "grad_norm": 0.5108677291834829,
+      "learning_rate": 4.673090866768412e-06,
+      "loss": 0.1791,
+      "step": 916
+    },
+    {
+      "epoch": 0.20660714788633228,
+      "grad_norm": 0.4859524561550817,
+      "learning_rate": 4.672169252337633e-06,
+      "loss": 0.165,
+      "step": 917
+    },
+    {
+      "epoch": 0.2068324555721407,
+      "grad_norm": 0.47506801577016344,
+      "learning_rate": 4.671246431782234e-06,
+      "loss": 0.1787,
+      "step": 918
+    },
+    {
+      "epoch": 0.20705776325794914,
+      "grad_norm": 0.4716251215451779,
+      "learning_rate": 4.670322405614621e-06,
+      "loss": 0.1676,
+      "step": 919
+    },
+    {
+      "epoch": 0.20728307094375756,
+      "grad_norm": 0.4888915064606123,
+      "learning_rate": 4.669397174347874e-06,
+      "loss": 0.1956,
+      "step": 920
+    },
+    {
+      "epoch": 0.207508378629566,
+      "grad_norm": 0.5008862880596165,
+      "learning_rate": 4.668470738495738e-06,
+      "loss": 0.1841,
+      "step": 921
+    },
+    {
+      "epoch": 0.20773368631537442,
+      "grad_norm": 0.4634482781420155,
+      "learning_rate": 4.667543098572627e-06,
+      "loss": 0.1839,
+      "step": 922
+    },
+    {
+      "epoch": 0.20795899400118287,
+      "grad_norm": 0.47429225006707093,
+      "learning_rate": 4.6666142550936286e-06,
+      "loss": 0.1634,
+      "step": 923
+    },
+    {
+      "epoch": 0.20818430168699129,
+      "grad_norm": 0.4661754900600514,
+      "learning_rate": 4.665684208574492e-06,
+      "loss": 0.1853,
+      "step": 924
+    },
+    {
+      "epoch": 0.20840960937279973,
+      "grad_norm": 0.46752723808840096,
+      "learning_rate": 4.664752959531638e-06,
+      "loss": 0.1731,
+      "step": 925
+    },
+    {
+      "epoch": 0.20863491705860815,
+      "grad_norm": 0.46142532792339114,
+      "learning_rate": 4.6638205084821544e-06,
+      "loss": 0.1707,
+      "step": 926
+    },
+    {
+      "epoch": 0.2088602247444166,
+      "grad_norm": 0.4891328446858097,
+      "learning_rate": 4.6628868559437964e-06,
+      "loss": 0.1763,
+      "step": 927
+    },
+    {
+      "epoch": 0.209085532430225,
+      "grad_norm": 0.5031128198962433,
+      "learning_rate": 4.661952002434988e-06,
+      "loss": 0.2023,
+      "step": 928
+    },
+    {
+      "epoch": 0.20931084011603346,
+      "grad_norm": 0.48714514401150427,
+      "learning_rate": 4.661015948474815e-06,
+      "loss": 0.1839,
+      "step": 929
+    },
+    {
+      "epoch": 0.20953614780184188,
+      "grad_norm": 0.4746123845118674,
+      "learning_rate": 4.660078694583037e-06,
+      "loss": 0.1782,
+      "step": 930
+    },
+    {
+      "epoch": 0.20976145548765032,
+      "grad_norm": 0.4664763084833289,
+      "learning_rate": 4.659140241280075e-06,
+      "loss": 0.1757,
+      "step": 931
+    },
+    {
+      "epoch": 0.20998676317345877,
+      "grad_norm": 0.4746618538561004,
+      "learning_rate": 4.658200589087016e-06,
+      "loss": 0.1776,
+      "step": 932
+    },
+    {
+      "epoch": 0.21021207085926719,
+      "grad_norm": 0.49077310248388495,
+      "learning_rate": 4.657259738525615e-06,
+      "loss": 0.1899,
+      "step": 933
+    },
+    {
+      "epoch": 0.21043737854507563,
+      "grad_norm": 0.4451674520587689,
+      "learning_rate": 4.656317690118291e-06,
+      "loss": 0.1683,
+      "step": 934
+    },
+    {
+      "epoch": 0.21066268623088405,
+      "grad_norm": 0.44999286347626516,
+      "learning_rate": 4.655374444388127e-06,
+      "loss": 0.1715,
+      "step": 935
+    },
+    {
+      "epoch": 0.2108879939166925,
+      "grad_norm": 0.5160001217026832,
+      "learning_rate": 4.654430001858874e-06,
+      "loss": 0.1885,
+      "step": 936
+    },
+    {
+      "epoch": 0.2111133016025009,
+      "grad_norm": 0.4842525872624328,
+      "learning_rate": 4.653484363054947e-06,
+      "loss": 0.1876,
+      "step": 937
+    },
+    {
+      "epoch": 0.21133860928830936,
+      "grad_norm": 0.49958988628695056,
+      "learning_rate": 4.6525375285014195e-06,
+      "loss": 0.1904,
+      "step": 938
+    },
+    {
+      "epoch": 0.21156391697411778,
+      "grad_norm": 0.4504957704587616,
+      "learning_rate": 4.651589498724037e-06,
+      "loss": 0.1727,
+      "step": 939
+    },
+    {
+      "epoch": 0.21178922465992622,
+      "grad_norm": 0.5015059809260454,
+      "learning_rate": 4.650640274249205e-06,
+      "loss": 0.1941,
+      "step": 940
+    },
+    {
+      "epoch": 0.21201453234573464,
+      "grad_norm": 0.476504641954251,
+      "learning_rate": 4.649689855603992e-06,
+      "loss": 0.175,
+      "step": 941
+    },
+    {
+      "epoch": 0.2122398400315431,
+      "grad_norm": 0.46840211368639195,
+      "learning_rate": 4.648738243316128e-06,
+      "loss": 0.1796,
+      "step": 942
+    },
+    {
+      "epoch": 0.2124651477173515,
+      "grad_norm": 0.4652139525522428,
+      "learning_rate": 4.647785437914011e-06,
+      "loss": 0.1748,
+      "step": 943
+    },
+    {
+      "epoch": 0.21269045540315995,
+      "grad_norm": 0.45886845605683524,
+      "learning_rate": 4.646831439926696e-06,
+      "loss": 0.1675,
+      "step": 944
+    },
+    {
+      "epoch": 0.21291576308896837,
+      "grad_norm": 0.47682867214434344,
+      "learning_rate": 4.645876249883903e-06,
+      "loss": 0.1786,
+      "step": 945
+    },
+    {
+      "epoch": 0.21314107077477681,
+      "grad_norm": 0.4452826345255325,
+      "learning_rate": 4.644919868316014e-06,
+      "loss": 0.1556,
+      "step": 946
+    },
+    {
+      "epoch": 0.21336637846058523,
+      "grad_norm": 0.5027037628671592,
+      "learning_rate": 4.643962295754073e-06,
+      "loss": 0.2023,
+      "step": 947
+    },
+    {
+      "epoch": 0.21359168614639368,
+      "grad_norm": 0.4838475880709884,
+      "learning_rate": 4.643003532729783e-06,
+      "loss": 0.1697,
+      "step": 948
+    },
+    {
+      "epoch": 0.2138169938322021,
+      "grad_norm": 0.4792563042333324,
+      "learning_rate": 4.642043579775509e-06,
+      "loss": 0.1774,
+      "step": 949
+    },
+    {
+      "epoch": 0.21404230151801054,
+      "grad_norm": 0.48434516547216877,
+      "learning_rate": 4.641082437424277e-06,
+      "loss": 0.1788,
+      "step": 950
+    },
+    {
+      "epoch": 0.21426760920381896,
+      "grad_norm": 0.4661730340891147,
+      "learning_rate": 4.640120106209776e-06,
+      "loss": 0.1753,
+      "step": 951
+    },
+    {
+      "epoch": 0.2144929168896274,
+      "grad_norm": 0.5180843258347066,
+      "learning_rate": 4.639156586666349e-06,
+      "loss": 0.1857,
+      "step": 952
+    },
+    {
+      "epoch": 0.21471822457543582,
+      "grad_norm": 0.514047193034028,
+      "learning_rate": 4.638191879329005e-06,
+      "loss": 0.1944,
+      "step": 953
+    },
+    {
+      "epoch": 0.21494353226124427,
+      "grad_norm": 0.4694837100451456,
+      "learning_rate": 4.63722598473341e-06,
+      "loss": 0.1686,
+      "step": 954
+    },
+    {
+      "epoch": 0.2151688399470527,
+      "grad_norm": 0.49442445420102776,
+      "learning_rate": 4.636258903415888e-06,
+      "loss": 0.1755,
+      "step": 955
+    },
+    {
+      "epoch": 0.21539414763286113,
+      "grad_norm": 0.48350627421275655,
+      "learning_rate": 4.635290635913425e-06,
+      "loss": 0.1772,
+      "step": 956
+    },
+    {
+      "epoch": 0.21561945531866955,
+      "grad_norm": 0.4694859909948316,
+      "learning_rate": 4.63432118276366e-06,
+      "loss": 0.1923,
+      "step": 957
+    },
+    {
+      "epoch": 0.215844763004478,
+      "grad_norm": 0.5416110255621837,
+      "learning_rate": 4.633350544504899e-06,
+      "loss": 0.179,
+      "step": 958
+    },
+    {
+      "epoch": 0.21607007069028641,
+      "grad_norm": 0.4465924442555772,
+      "learning_rate": 4.632378721676098e-06,
+      "loss": 0.1742,
+      "step": 959
+    },
+    {
+      "epoch": 0.21629537837609486,
+      "grad_norm": 0.458481081744707,
+      "learning_rate": 4.6314057148168765e-06,
+      "loss": 0.1745,
+      "step": 960
+    },
+    {
+      "epoch": 0.21652068606190328,
+      "grad_norm": 0.4547664974315061,
+      "learning_rate": 4.6304315244675065e-06,
+      "loss": 0.1774,
+      "step": 961
+    },
+    {
+      "epoch": 0.21674599374771172,
+      "grad_norm": 0.5000337392694555,
+      "learning_rate": 4.629456151168921e-06,
+      "loss": 0.1848,
+      "step": 962
+    },
+    {
+      "epoch": 0.21697130143352014,
+      "grad_norm": 0.45255086182737697,
+      "learning_rate": 4.628479595462708e-06,
+      "loss": 0.1767,
+      "step": 963
+    },
+    {
+      "epoch": 0.2171966091193286,
+      "grad_norm": 0.5050866405244006,
+      "learning_rate": 4.627501857891113e-06,
+      "loss": 0.1794,
+      "step": 964
+    },
+    {
+      "epoch": 0.217421916805137,
+      "grad_norm": 0.5046057153343321,
+      "learning_rate": 4.626522938997037e-06,
+      "loss": 0.1901,
+      "step": 965
+    },
+    {
+      "epoch": 0.21764722449094545,
+      "grad_norm": 0.49654858884976455,
+      "learning_rate": 4.625542839324036e-06,
+      "loss": 0.1896,
+      "step": 966
+    },
+    {
+      "epoch": 0.21787253217675387,
+      "grad_norm": 0.4619603328007907,
+      "learning_rate": 4.624561559416324e-06,
+      "loss": 0.1722,
+      "step": 967
+    },
+    {
+      "epoch": 0.21809783986256231,
+      "grad_norm": 0.4714237924999947,
+      "learning_rate": 4.623579099818769e-06,
+      "loss": 0.1811,
+      "step": 968
+    },
+    {
+      "epoch": 0.21832314754837073,
+      "grad_norm": 0.4727906799742656,
+      "learning_rate": 4.6225954610768945e-06,
+      "loss": 0.1641,
+      "step": 969
+    },
+    {
+      "epoch": 0.21854845523417918,
+      "grad_norm": 0.4988859251827525,
+      "learning_rate": 4.621610643736878e-06,
+      "loss": 0.1807,
+      "step": 970
+    },
+    {
+      "epoch": 0.2187737629199876,
+      "grad_norm": 0.5060976705176674,
+      "learning_rate": 4.620624648345552e-06,
+      "loss": 0.1749,
+      "step": 971
+    },
+    {
+      "epoch": 0.21899907060579604,
+      "grad_norm": 0.47793522949791656,
+      "learning_rate": 4.6196374754504024e-06,
+      "loss": 0.1826,
+      "step": 972
+    },
+    {
+      "epoch": 0.21922437829160446,
+      "grad_norm": 0.5033809177523444,
+      "learning_rate": 4.61864912559957e-06,
+      "loss": 0.1944,
+      "step": 973
+    },
+    {
+      "epoch": 0.2194496859774129,
+      "grad_norm": 0.45111023806215533,
+      "learning_rate": 4.617659599341849e-06,
+      "loss": 0.1621,
+      "step": 974
+    },
+    {
+      "epoch": 0.21967499366322132,
+      "grad_norm": 0.548417164765913,
+      "learning_rate": 4.616668897226686e-06,
+      "loss": 0.1767,
+      "step": 975
+    },
+    {
+      "epoch": 0.21990030134902977,
+      "grad_norm": 0.5226621764784928,
+      "learning_rate": 4.615677019804182e-06,
+      "loss": 0.1796,
+      "step": 976
+    },
+    {
+      "epoch": 0.2201256090348382,
+      "grad_norm": 0.46896809226111524,
+      "learning_rate": 4.6146839676250875e-06,
+      "loss": 0.1679,
+      "step": 977
+    },
+    {
+      "epoch": 0.22035091672064663,
+      "grad_norm": 0.4889116397075663,
+      "learning_rate": 4.6136897412408084e-06,
+      "loss": 0.1894,
+      "step": 978
+    },
+    {
+      "epoch": 0.22057622440645505,
+      "grad_norm": 0.45531634545026695,
+      "learning_rate": 4.612694341203403e-06,
+      "loss": 0.1757,
+      "step": 979
+    },
+    {
+      "epoch": 0.2208015320922635,
+      "grad_norm": 0.5139339558980113,
+      "learning_rate": 4.611697768065577e-06,
+      "loss": 0.1963,
+      "step": 980
+    },
+    {
+      "epoch": 0.22102683977807194,
+      "grad_norm": 0.461006150958274,
+      "learning_rate": 4.610700022380692e-06,
+      "loss": 0.174,
+      "step": 981
+    },
+    {
+      "epoch": 0.22125214746388036,
+      "grad_norm": 0.4468605106113933,
+      "learning_rate": 4.609701104702759e-06,
+      "loss": 0.1587,
+      "step": 982
+    },
+    {
+      "epoch": 0.2214774551496888,
+      "grad_norm": 0.43419083405156583,
+      "learning_rate": 4.6087010155864394e-06,
+      "loss": 0.1551,
+      "step": 983
+    },
+    {
+      "epoch": 0.22170276283549722,
+      "grad_norm": 0.48688751814505576,
+      "learning_rate": 4.607699755587046e-06,
+      "loss": 0.1872,
+      "step": 984
+    },
+    {
+      "epoch": 0.22192807052130567,
+      "grad_norm": 0.49965055488304555,
+      "learning_rate": 4.60669732526054e-06,
+      "loss": 0.1955,
+      "step": 985
+    },
+    {
+      "epoch": 0.2221533782071141,
+      "grad_norm": 0.4834893714431329,
+      "learning_rate": 4.605693725163536e-06,
+      "loss": 0.1856,
+      "step": 986
+    },
+    {
+      "epoch": 0.22237868589292253,
+      "grad_norm": 0.4990375863342495,
+      "learning_rate": 4.6046889558532925e-06,
+      "loss": 0.18,
+      "step": 987
+    },
+    {
+      "epoch": 0.22260399357873095,
+      "grad_norm": 0.484244316222459,
+      "learning_rate": 4.603683017887722e-06,
+      "loss": 0.1725,
+      "step": 988
+    },
+    {
+      "epoch": 0.2228293012645394,
+      "grad_norm": 0.467849362229871,
+      "learning_rate": 4.602675911825386e-06,
+      "loss": 0.172,
+      "step": 989
+    },
+    {
+      "epoch": 0.22305460895034782,
+      "grad_norm": 0.46897024187918585,
+      "learning_rate": 4.6016676382254895e-06,
+      "loss": 0.1688,
+      "step": 990
+    },
+    {
+      "epoch": 0.22327991663615626,
+      "grad_norm": 0.4773396813265902,
+      "learning_rate": 4.600658197647892e-06,
+      "loss": 0.1685,
+      "step": 991
+    },
+    {
+      "epoch": 0.22350522432196468,
+      "grad_norm": 0.5064121299266291,
+      "learning_rate": 4.5996475906530955e-06,
+      "loss": 0.1806,
+      "step": 992
+    },
+    {
+      "epoch": 0.22373053200777313,
+      "grad_norm": 0.47765732789338444,
+      "learning_rate": 4.598635817802256e-06,
+      "loss": 0.1718,
+      "step": 993
+    },
+    {
+      "epoch": 0.22395583969358154,
+      "grad_norm": 0.5291926414513833,
+      "learning_rate": 4.597622879657171e-06,
+      "loss": 0.1725,
+      "step": 994
+    },
+    {
+      "epoch": 0.22418114737939,
+      "grad_norm": 0.4797571989600491,
+      "learning_rate": 4.596608776780287e-06,
+      "loss": 0.1706,
+      "step": 995
+    },
+    {
+      "epoch": 0.2244064550651984,
+      "grad_norm": 0.4864708362280325,
+      "learning_rate": 4.595593509734699e-06,
+      "loss": 0.1766,
+      "step": 996
+    },
+    {
+      "epoch": 0.22463176275100685,
+      "grad_norm": 0.45907950569959,
+      "learning_rate": 4.594577079084146e-06,
+      "loss": 0.1674,
+      "step": 997
+    },
+    {
+      "epoch": 0.22485707043681527,
+      "grad_norm": 0.5055242979755723,
+      "learning_rate": 4.593559485393015e-06,
+      "loss": 0.1752,
+      "step": 998
+    },
+    {
+      "epoch": 0.22508237812262372,
+      "grad_norm": 0.5001209119984309,
+      "learning_rate": 4.592540729226336e-06,
+      "loss": 0.1821,
+      "step": 999
+    },
+    {
+      "epoch": 0.22530768580843213,
+      "grad_norm": 0.5040109793881857,
+      "learning_rate": 4.591520811149787e-06,
+      "loss": 0.1785,
+      "step": 1000
+    },
+    {
+      "epoch": 0.22530768580843213,
+      "eval_loss": 0.1766696572303772,
+      "eval_runtime": 57.3568,
+      "eval_samples_per_second": 50.038,
+      "eval_steps_per_second": 6.259,
+      "step": 1000
+    },
+    {
+      "epoch": 0.22553299349424058,
+      "grad_norm": 0.48462064594502124,
+      "learning_rate": 4.590499731729692e-06,
+      "loss": 0.1686,
+      "step": 1001
+    },
+    {
+      "epoch": 0.225758301180049,
+      "grad_norm": 0.4601104950424043,
+      "learning_rate": 4.589477491533016e-06,
+      "loss": 0.1556,
+      "step": 1002
+    },
+    {
+      "epoch": 0.22598360886585744,
+      "grad_norm": 0.49464098877745066,
+      "learning_rate": 4.588454091127373e-06,
+      "loss": 0.1796,
+      "step": 1003
+    },
+    {
+      "epoch": 0.22620891655166586,
+      "grad_norm": 0.5148232425969892,
+      "learning_rate": 4.587429531081019e-06,
+      "loss": 0.1914,
+      "step": 1004
+    },
+    {
+      "epoch": 0.2264342242374743,
+      "grad_norm": 0.47731591236053894,
+      "learning_rate": 4.586403811962852e-06,
+      "loss": 0.1604,
+      "step": 1005
+    },
+    {
+      "epoch": 0.22665953192328273,
+      "grad_norm": 0.4551856162700306,
+      "learning_rate": 4.585376934342418e-06,
+      "loss": 0.1641,
+      "step": 1006
+    },
+    {
+      "epoch": 0.22688483960909117,
+      "grad_norm": 0.4637388238429355,
+      "learning_rate": 4.584348898789901e-06,
+      "loss": 0.1695,
+      "step": 1007
+    },
+    {
+      "epoch": 0.2271101472948996,
+      "grad_norm": 0.46640928542419985,
+      "learning_rate": 4.583319705876133e-06,
+      "loss": 0.1778,
+      "step": 1008
+    },
+    {
+      "epoch": 0.22733545498070803,
+      "grad_norm": 0.4910752180840778,
+      "learning_rate": 4.5822893561725864e-06,
+      "loss": 0.1709,
+      "step": 1009
+    },
+    {
+      "epoch": 0.22756076266651645,
+      "grad_norm": 0.4975885196146549,
+      "learning_rate": 4.581257850251376e-06,
+      "loss": 0.1906,
+      "step": 1010
+    },
+    {
+      "epoch": 0.2277860703523249,
+      "grad_norm": 0.48897736900788225,
+      "learning_rate": 4.580225188685257e-06,
+      "loss": 0.1707,
+      "step": 1011
+    },
+    {
+      "epoch": 0.22801137803813332,
+      "grad_norm": 0.43130461495166755,
+      "learning_rate": 4.579191372047631e-06,
+      "loss": 0.1618,
+      "step": 1012
+    },
+    {
+      "epoch": 0.22823668572394176,
+      "grad_norm": 0.5346355301760246,
+      "learning_rate": 4.578156400912535e-06,
+      "loss": 0.2052,
+      "step": 1013
+    },
+    {
+      "epoch": 0.22846199340975018,
+      "grad_norm": 0.5120793734577033,
+      "learning_rate": 4.577120275854649e-06,
+      "loss": 0.1786,
+      "step": 1014
+    },
+    {
+      "epoch": 0.22868730109555863,
+      "grad_norm": 0.47045662878311806,
+      "learning_rate": 4.576082997449298e-06,
+      "loss": 0.1801,
+      "step": 1015
+    },
+    {
+      "epoch": 0.22891260878136704,
+      "grad_norm": 0.4897903682435324,
+      "learning_rate": 4.5750445662724426e-06,
+      "loss": 0.182,
+      "step": 1016
+    },
+    {
+      "epoch": 0.2291379164671755,
+      "grad_norm": 0.4496213951131232,
+      "learning_rate": 4.574004982900684e-06,
+      "loss": 0.1726,
+      "step": 1017
+    },
+    {
+      "epoch": 0.2293632241529839,
+      "grad_norm": 0.4796218909713557,
+      "learning_rate": 4.572964247911265e-06,
+      "loss": 0.1806,
+      "step": 1018
+    },
+    {
+      "epoch": 0.22958853183879235,
+      "grad_norm": 0.48130542448321,
+      "learning_rate": 4.5719223618820666e-06,
+      "loss": 0.1884,
+      "step": 1019
+    },
+    {
+      "epoch": 0.22981383952460077,
+      "grad_norm": 0.46705890705030956,
+      "learning_rate": 4.5708793253916104e-06,
+      "loss": 0.171,
+      "step": 1020
+    },
+    {
+      "epoch": 0.23003914721040922,
+      "grad_norm": 0.4598433200960336,
+      "learning_rate": 4.569835139019054e-06,
+      "loss": 0.1625,
+      "step": 1021
+    },
+    {
+      "epoch": 0.23026445489621764,
+      "grad_norm": 0.4798183615941154,
+      "learning_rate": 4.568789803344196e-06,
+      "loss": 0.1854,
+      "step": 1022
+    },
+    {
+      "epoch": 0.23048976258202608,
+      "grad_norm": 0.5049857914668311,
+      "learning_rate": 4.567743318947472e-06,
+      "loss": 0.1723,
+      "step": 1023
+    },
+    {
+      "epoch": 0.2307150702678345,
+      "grad_norm": 0.5133377849430174,
+      "learning_rate": 4.566695686409957e-06,
+      "loss": 0.1969,
+      "step": 1024
+    },
+    {
+      "epoch": 0.23094037795364294,
+      "grad_norm": 0.4700007173190981,
+      "learning_rate": 4.56564690631336e-06,
+      "loss": 0.175,
+      "step": 1025
+    },
+    {
+      "epoch": 0.23116568563945136,
+      "grad_norm": 0.47259284661188733,
+      "learning_rate": 4.564596979240031e-06,
+      "loss": 0.1648,
+      "step": 1026
+    },
+    {
+      "epoch": 0.2313909933252598,
+      "grad_norm": 0.46806795756718333,
+      "learning_rate": 4.563545905772956e-06,
+      "loss": 0.1684,
+      "step": 1027
+    },
+    {
+      "epoch": 0.23161630101106823,
+      "grad_norm": 0.45862602476089215,
+      "learning_rate": 4.562493686495756e-06,
+      "loss": 0.1684,
+      "step": 1028
+    },
+    {
+      "epoch": 0.23184160869687667,
+      "grad_norm": 0.4633999365729002,
+      "learning_rate": 4.56144032199269e-06,
+      "loss": 0.1586,
+      "step": 1029
+    },
+    {
+      "epoch": 0.23206691638268512,
+      "grad_norm": 0.46864023859664566,
+      "learning_rate": 4.56038581284865e-06,
+      "loss": 0.167,
+      "step": 1030
+    },
+    {
+      "epoch": 0.23229222406849354,
+      "grad_norm": 0.4565233615702026,
+      "learning_rate": 4.559330159649166e-06,
+      "loss": 0.1701,
+      "step": 1031
+    },
+    {
+      "epoch": 0.23251753175430198,
+      "grad_norm": 0.46694580167435407,
+      "learning_rate": 4.558273362980406e-06,
+      "loss": 0.1682,
+      "step": 1032
+    },
+    {
+      "epoch": 0.2327428394401104,
+      "grad_norm": 0.448802180941964,
+      "learning_rate": 4.557215423429167e-06,
+      "loss": 0.1692,
+      "step": 1033
+    },
+    {
+      "epoch": 0.23296814712591885,
+      "grad_norm": 0.46242148368245545,
+      "learning_rate": 4.556156341582884e-06,
+      "loss": 0.1757,
+      "step": 1034
+    },
+    {
+      "epoch": 0.23319345481172726,
+      "grad_norm": 0.4702960441016916,
+      "learning_rate": 4.555096118029625e-06,
+      "loss": 0.1649,
+      "step": 1035
+    },
+    {
+      "epoch": 0.2334187624975357,
+      "grad_norm": 0.4603866097849357,
+      "learning_rate": 4.5540347533580935e-06,
+      "loss": 0.1743,
+      "step": 1036
+    },
+    {
+      "epoch": 0.23364407018334413,
+      "grad_norm": 0.5016444270737936,
+      "learning_rate": 4.5529722481576265e-06,
+      "loss": 0.1659,
+      "step": 1037
+    },
+    {
+      "epoch": 0.23386937786915257,
+      "grad_norm": 0.46197661033483206,
+      "learning_rate": 4.551908603018191e-06,
+      "loss": 0.1718,
+      "step": 1038
+    },
+    {
+      "epoch": 0.234094685554961,
+      "grad_norm": 0.4576051232314433,
+      "learning_rate": 4.550843818530392e-06,
+      "loss": 0.1754,
+      "step": 1039
+    },
+    {
+      "epoch": 0.23431999324076944,
+      "grad_norm": 0.48921900894809295,
+      "learning_rate": 4.549777895285464e-06,
+      "loss": 0.1681,
+      "step": 1040
+    },
+    {
+      "epoch": 0.23454530092657785,
+      "grad_norm": 0.4969985437197727,
+      "learning_rate": 4.548710833875273e-06,
+      "loss": 0.1776,
+      "step": 1041
+    },
+    {
+      "epoch": 0.2347706086123863,
+      "grad_norm": 0.4926942037575357,
+      "learning_rate": 4.547642634892321e-06,
+      "loss": 0.1711,
+      "step": 1042
+    },
+    {
+      "epoch": 0.23499591629819472,
+      "grad_norm": 0.5099757228479024,
+      "learning_rate": 4.5465732989297365e-06,
+      "loss": 0.169,
+      "step": 1043
+    },
+    {
+      "epoch": 0.23522122398400316,
+      "grad_norm": 0.4941141847733085,
+      "learning_rate": 4.545502826581284e-06,
+      "loss": 0.1758,
+      "step": 1044
+    },
+    {
+      "epoch": 0.23544653166981158,
+      "grad_norm": 0.4992853772276655,
+      "learning_rate": 4.5444312184413554e-06,
+      "loss": 0.1896,
+      "step": 1045
+    },
+    {
+      "epoch": 0.23567183935562003,
+      "grad_norm": 0.5077593719343416,
+      "learning_rate": 4.543358475104975e-06,
+      "loss": 0.1731,
+      "step": 1046
+    },
+    {
+      "epoch": 0.23589714704142845,
+      "grad_norm": 0.432534135028472,
+      "learning_rate": 4.5422845971677985e-06,
+      "loss": 0.1606,
+      "step": 1047
+    },
+    {
+      "epoch": 0.2361224547272369,
+      "grad_norm": 0.46569122347590924,
+      "learning_rate": 4.541209585226109e-06,
+      "loss": 0.1645,
+      "step": 1048
+    },
+    {
+      "epoch": 0.2363477624130453,
+      "grad_norm": 0.4851233704311572,
+      "learning_rate": 4.5401334398768195e-06,
+      "loss": 0.164,
+      "step": 1049
+    },
+    {
+      "epoch": 0.23657307009885375,
+      "grad_norm": 0.49161973544017507,
+      "learning_rate": 4.539056161717477e-06,
+      "loss": 0.1704,
+      "step": 1050
+    },
+    {
+      "epoch": 0.23679837778466217,
+      "grad_norm": 0.4705002761219879,
+      "learning_rate": 4.53797775134625e-06,
+      "loss": 0.1572,
+      "step": 1051
+    },
+    {
+      "epoch": 0.23702368547047062,
+      "grad_norm": 0.5074699444832521,
+      "learning_rate": 4.536898209361942e-06,
+      "loss": 0.1817,
+      "step": 1052
+    },
+    {
+      "epoch": 0.23724899315627904,
+      "grad_norm": 0.4753176300220157,
+      "learning_rate": 4.535817536363981e-06,
+      "loss": 0.1679,
+      "step": 1053
+    },
+    {
+      "epoch": 0.23747430084208748,
+      "grad_norm": 0.45572087945610384,
+      "learning_rate": 4.5347357329524254e-06,
+      "loss": 0.1557,
+      "step": 1054
+    },
+    {
+      "epoch": 0.2376996085278959,
+      "grad_norm": 0.47954801189962154,
+      "learning_rate": 4.53365279972796e-06,
+      "loss": 0.1739,
+      "step": 1055
+    },
+    {
+      "epoch": 0.23792491621370435,
+      "grad_norm": 0.4936922163351292,
+      "learning_rate": 4.532568737291898e-06,
+      "loss": 0.1708,
+      "step": 1056
+    },
+    {
+      "epoch": 0.23815022389951276,
+      "grad_norm": 0.487882288980118,
+      "learning_rate": 4.531483546246177e-06,
+      "loss": 0.1867,
+      "step": 1057
+    },
+    {
+      "epoch": 0.2383755315853212,
+      "grad_norm": 0.5022570562698242,
+      "learning_rate": 4.530397227193365e-06,
+      "loss": 0.1763,
+      "step": 1058
+    },
+    {
+      "epoch": 0.23860083927112963,
+      "grad_norm": 0.5098565893687929,
+      "learning_rate": 4.529309780736654e-06,
+      "loss": 0.1821,
+      "step": 1059
+    },
+    {
+      "epoch": 0.23882614695693807,
+      "grad_norm": 0.5471270750459719,
+      "learning_rate": 4.528221207479862e-06,
+      "loss": 0.1897,
+      "step": 1060
+    },
+    {
+      "epoch": 0.2390514546427465,
+      "grad_norm": 0.46289499595120953,
+      "learning_rate": 4.527131508027433e-06,
+      "loss": 0.1579,
+      "step": 1061
+    },
+    {
+      "epoch": 0.23927676232855494,
+      "grad_norm": 0.49338984383186363,
+      "learning_rate": 4.5260406829844364e-06,
+      "loss": 0.1807,
+      "step": 1062
+    },
+    {
+      "epoch": 0.23950207001436336,
+      "grad_norm": 0.5029016610935466,
+      "learning_rate": 4.524948732956568e-06,
+      "loss": 0.1948,
+      "step": 1063
+    },
+    {
+      "epoch": 0.2397273777001718,
+      "grad_norm": 0.49997268202495476,
+      "learning_rate": 4.523855658550146e-06,
+      "loss": 0.1792,
+      "step": 1064
+    },
+    {
+      "epoch": 0.23995268538598022,
+      "grad_norm": 0.48933115982084735,
+      "learning_rate": 4.522761460372114e-06,
+      "loss": 0.178,
+      "step": 1065
+    },
+    {
+      "epoch": 0.24017799307178866,
+      "grad_norm": 0.47946740350317474,
+      "learning_rate": 4.521666139030039e-06,
+      "loss": 0.1677,
+      "step": 1066
+    },
+    {
+      "epoch": 0.24040330075759708,
+      "grad_norm": 0.4184658820157611,
+      "learning_rate": 4.520569695132113e-06,
+      "loss": 0.15,
+      "step": 1067
+    },
+    {
+      "epoch": 0.24062860844340553,
+      "grad_norm": 0.5197049444095906,
+      "learning_rate": 4.51947212928715e-06,
+      "loss": 0.1794,
+      "step": 1068
+    },
+    {
+      "epoch": 0.24085391612921395,
+      "grad_norm": 0.5201746421812771,
+      "learning_rate": 4.518373442104587e-06,
+      "loss": 0.1807,
+      "step": 1069
+    },
+    {
+      "epoch": 0.2410792238150224,
+      "grad_norm": 0.4459522405235029,
+      "learning_rate": 4.5172736341944845e-06,
+      "loss": 0.1659,
+      "step": 1070
+    },
+    {
+      "epoch": 0.2413045315008308,
+      "grad_norm": 0.4914213660575991,
+      "learning_rate": 4.516172706167525e-06,
+      "loss": 0.1759,
+      "step": 1071
+    },
+    {
+      "epoch": 0.24152983918663926,
+      "grad_norm": 0.47911999094848545,
+      "learning_rate": 4.515070658635013e-06,
+      "loss": 0.1768,
+      "step": 1072
+    },
+    {
+      "epoch": 0.24175514687244767,
+      "grad_norm": 0.4892612077155755,
+      "learning_rate": 4.513967492208874e-06,
+      "loss": 0.1715,
+      "step": 1073
+    },
+    {
+      "epoch": 0.24198045455825612,
+      "grad_norm": 0.5259185134683907,
+      "learning_rate": 4.512863207501654e-06,
+      "loss": 0.1916,
+      "step": 1074
+    },
+    {
+      "epoch": 0.24220576224406454,
+      "grad_norm": 0.459465389299004,
+      "learning_rate": 4.511757805126523e-06,
+      "loss": 0.172,
+      "step": 1075
+    },
+    {
+      "epoch": 0.24243106992987298,
+      "grad_norm": 0.47910539989958506,
+      "learning_rate": 4.510651285697269e-06,
+      "loss": 0.1726,
+      "step": 1076
+    },
+    {
+      "epoch": 0.2426563776156814,
+      "grad_norm": 0.4701014911994522,
+      "learning_rate": 4.509543649828302e-06,
+      "loss": 0.1681,
+      "step": 1077
+    },
+    {
+      "epoch": 0.24288168530148985,
+      "grad_norm": 0.46213489937923835,
+      "learning_rate": 4.5084348981346495e-06,
+      "loss": 0.1719,
+      "step": 1078
+    },
+    {
+      "epoch": 0.2431069929872983,
+      "grad_norm": 0.4935132048768077,
+      "learning_rate": 4.507325031231959e-06,
+      "loss": 0.1825,
+      "step": 1079
+    },
+    {
+      "epoch": 0.2433323006731067,
+      "grad_norm": 0.48392629751946686,
+      "learning_rate": 4.506214049736502e-06,
+      "loss": 0.1713,
+      "step": 1080
+    },
+    {
+      "epoch": 0.24355760835891516,
+      "grad_norm": 0.5080595176990748,
+      "learning_rate": 4.505101954265161e-06,
+      "loss": 0.1903,
+      "step": 1081
+    },
+    {
+      "epoch": 0.24378291604472357,
+      "grad_norm": 0.4673535025076435,
+      "learning_rate": 4.503988745435443e-06,
+      "loss": 0.1608,
+      "step": 1082
+    },
+    {
+      "epoch": 0.24400822373053202,
+      "grad_norm": 0.4821960030125238,
+      "learning_rate": 4.502874423865473e-06,
+      "loss": 0.1817,
+      "step": 1083
+    },
+    {
+      "epoch": 0.24423353141634044,
+      "grad_norm": 0.4825223552549021,
+      "learning_rate": 4.5017589901739885e-06,
+      "loss": 0.1703,
+      "step": 1084
+    },
+    {
+      "epoch": 0.24445883910214888,
+      "grad_norm": 0.49338394902001637,
+      "learning_rate": 4.500642444980352e-06,
+      "loss": 0.1737,
+      "step": 1085
+    },
+    {
+      "epoch": 0.2446841467879573,
+      "grad_norm": 0.4545645277155261,
+      "learning_rate": 4.499524788904537e-06,
+      "loss": 0.1636,
+      "step": 1086
+    },
+    {
+      "epoch": 0.24490945447376575,
+      "grad_norm": 0.48306624324737196,
+      "learning_rate": 4.498406022567137e-06,
+      "loss": 0.1765,
+      "step": 1087
+    },
+    {
+      "epoch": 0.24513476215957417,
+      "grad_norm": 0.4749043885580964,
+      "learning_rate": 4.497286146589361e-06,
+      "loss": 0.1693,
+      "step": 1088
+    },
+    {
+      "epoch": 0.2453600698453826,
+      "grad_norm": 0.4712090941375913,
+      "learning_rate": 4.4961651615930344e-06,
+      "loss": 0.1647,
+      "step": 1089
+    },
+    {
+      "epoch": 0.24558537753119103,
+      "grad_norm": 0.48622341817676057,
+      "learning_rate": 4.4950430682005995e-06,
+      "loss": 0.1672,
+      "step": 1090
+    },
+    {
+      "epoch": 0.24581068521699947,
+      "grad_norm": 0.46188635009124457,
+      "learning_rate": 4.493919867035112e-06,
+      "loss": 0.1477,
+      "step": 1091
+    },
+    {
+      "epoch": 0.2460359929028079,
+      "grad_norm": 0.4548856414210293,
+      "learning_rate": 4.492795558720242e-06,
+      "loss": 0.1578,
+      "step": 1092
+    },
+    {
+      "epoch": 0.24626130058861634,
+      "grad_norm": 0.46865916685520176,
+      "learning_rate": 4.491670143880279e-06,
+      "loss": 0.1577,
+      "step": 1093
+    },
+    {
+      "epoch": 0.24648660827442476,
+      "grad_norm": 0.47334725360934393,
+      "learning_rate": 4.490543623140123e-06,
+      "loss": 0.1785,
+      "step": 1094
+    },
+    {
+      "epoch": 0.2467119159602332,
+      "grad_norm": 0.4903085986965832,
+      "learning_rate": 4.489415997125288e-06,
+      "loss": 0.1787,
+      "step": 1095
+    },
+    {
+      "epoch": 0.24693722364604162,
+      "grad_norm": 0.4987257674352878,
+      "learning_rate": 4.488287266461904e-06,
+      "loss": 0.1822,
+      "step": 1096
+    },
+    {
+      "epoch": 0.24716253133185007,
+      "grad_norm": 0.49514915506271256,
+      "learning_rate": 4.487157431776712e-06,
+      "loss": 0.1786,
+      "step": 1097
+    },
+    {
+      "epoch": 0.24738783901765848,
+      "grad_norm": 0.4895397147505216,
+      "learning_rate": 4.486026493697067e-06,
+      "loss": 0.176,
+      "step": 1098
+    },
+    {
+      "epoch": 0.24761314670346693,
+      "grad_norm": 0.4750865000533062,
+      "learning_rate": 4.484894452850937e-06,
+      "loss": 0.1662,
+      "step": 1099
+    },
+    {
+      "epoch": 0.24783845438927535,
+      "grad_norm": 0.5231586359914755,
+      "learning_rate": 4.483761309866902e-06,
+      "loss": 0.1932,
+      "step": 1100
+    },
+    {
+      "epoch": 0.2480637620750838,
+      "grad_norm": 0.49524527920469363,
+      "learning_rate": 4.482627065374155e-06,
+      "loss": 0.1854,
+      "step": 1101
+    },
+    {
+      "epoch": 0.2482890697608922,
+      "grad_norm": 0.4674761617417515,
+      "learning_rate": 4.481491720002499e-06,
+      "loss": 0.162,
+      "step": 1102
+    },
+    {
+      "epoch": 0.24851437744670066,
+      "grad_norm": 0.49289841749272967,
+      "learning_rate": 4.4803552743823495e-06,
+      "loss": 0.1754,
+      "step": 1103
+    },
+    {
+      "epoch": 0.24873968513250908,
+      "grad_norm": 0.5096228946461686,
+      "learning_rate": 4.479217729144731e-06,
+      "loss": 0.1813,
+      "step": 1104
+    },
+    {
+      "epoch": 0.24896499281831752,
+      "grad_norm": 0.4709558119876889,
+      "learning_rate": 4.478079084921282e-06,
+      "loss": 0.159,
+      "step": 1105
+    },
+    {
+      "epoch": 0.24919030050412594,
+      "grad_norm": 0.46290162539634433,
+      "learning_rate": 4.476939342344246e-06,
+      "loss": 0.163,
+      "step": 1106
+    },
+    {
+      "epoch": 0.24941560818993438,
+      "grad_norm": 0.47236146938099277,
+      "learning_rate": 4.475798502046484e-06,
+      "loss": 0.1705,
+      "step": 1107
+    },
+    {
+      "epoch": 0.2496409158757428,
+      "grad_norm": 0.4759949192730458,
+      "learning_rate": 4.474656564661458e-06,
+      "loss": 0.1652,
+      "step": 1108
+    },
+    {
+      "epoch": 0.24986622356155125,
+      "grad_norm": 0.49013257055563464,
+      "learning_rate": 4.473513530823246e-06,
+      "loss": 0.1851,
+      "step": 1109
+    },
+    {
+      "epoch": 0.25009153124735967,
+      "grad_norm": 0.46874665727501796,
+      "learning_rate": 4.472369401166531e-06,
+      "loss": 0.1724,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2503168389331681,
+      "grad_norm": 0.4621201673920498,
+      "learning_rate": 4.471224176326605e-06,
+      "loss": 0.1675,
+      "step": 1111
+    },
+    {
+      "epoch": 0.25054214661897656,
+      "grad_norm": 0.45656340706343534,
+      "learning_rate": 4.47007785693937e-06,
+      "loss": 0.1751,
+      "step": 1112
+    },
+    {
+      "epoch": 0.250767454304785,
+      "grad_norm": 0.48141602205933126,
+      "learning_rate": 4.468930443641333e-06,
+      "loss": 0.1751,
+      "step": 1113
+    },
+    {
+      "epoch": 0.2509927619905934,
+      "grad_norm": 0.4858128449295977,
+      "learning_rate": 4.467781937069611e-06,
+      "loss": 0.1775,
+      "step": 1114
+    },
+    {
+      "epoch": 0.2512180696764018,
+      "grad_norm": 0.49393776518442545,
+      "learning_rate": 4.466632337861926e-06,
+      "loss": 0.1693,
+      "step": 1115
+    },
+    {
+      "epoch": 0.2514433773622103,
+      "grad_norm": 0.5125458797038791,
+      "learning_rate": 4.465481646656608e-06,
+      "loss": 0.1867,
+      "step": 1116
+    },
+    {
+      "epoch": 0.2516686850480187,
+      "grad_norm": 0.475803038903292,
+      "learning_rate": 4.464329864092593e-06,
+      "loss": 0.1716,
+      "step": 1117
+    },
+    {
+      "epoch": 0.2518939927338271,
+      "grad_norm": 0.5419619115552078,
+      "learning_rate": 4.463176990809423e-06,
+      "loss": 0.1828,
+      "step": 1118
+    },
+    {
+      "epoch": 0.25211930041963554,
+      "grad_norm": 0.4639433198703533,
+      "learning_rate": 4.462023027447246e-06,
+      "loss": 0.1723,
+      "step": 1119
+    },
+    {
+      "epoch": 0.252344608105444,
+      "grad_norm": 0.46793145441851886,
+      "learning_rate": 4.460867974646814e-06,
+      "loss": 0.1629,
+      "step": 1120
+    },
+    {
+      "epoch": 0.25256991579125243,
+      "grad_norm": 0.5411894292034929,
+      "learning_rate": 4.459711833049485e-06,
+      "loss": 0.1961,
+      "step": 1121
+    },
+    {
+      "epoch": 0.25279522347706085,
+      "grad_norm": 0.4989548751939043,
+      "learning_rate": 4.45855460329722e-06,
+      "loss": 0.1731,
+      "step": 1122
+    },
+    {
+      "epoch": 0.25302053116286927,
+      "grad_norm": 0.5243782674199724,
+      "learning_rate": 4.457396286032589e-06,
+      "loss": 0.1776,
+      "step": 1123
+    },
+    {
+      "epoch": 0.25324583884867774,
+      "grad_norm": 0.4886069442128809,
+      "learning_rate": 4.45623688189876e-06,
+      "loss": 0.1707,
+      "step": 1124
+    },
+    {
+      "epoch": 0.25347114653448616,
+      "grad_norm": 0.5012040772255312,
+      "learning_rate": 4.455076391539507e-06,
+      "loss": 0.1812,
+      "step": 1125
+    },
+    {
+      "epoch": 0.2536964542202946,
+      "grad_norm": 0.571500324082683,
+      "learning_rate": 4.453914815599206e-06,
+      "loss": 0.1659,
+      "step": 1126
+    },
+    {
+      "epoch": 0.25392176190610305,
+      "grad_norm": 0.4948654591758799,
+      "learning_rate": 4.45275215472284e-06,
+      "loss": 0.1757,
+      "step": 1127
+    },
+    {
+      "epoch": 0.25414706959191147,
+      "grad_norm": 0.4544375354575435,
+      "learning_rate": 4.451588409555988e-06,
+      "loss": 0.162,
+      "step": 1128
+    },
+    {
+      "epoch": 0.2543723772777199,
+      "grad_norm": 0.46357726308267455,
+      "learning_rate": 4.450423580744837e-06,
+      "loss": 0.1668,
+      "step": 1129
+    },
+    {
+      "epoch": 0.2545976849635283,
+      "grad_norm": 0.4820056825601605,
+      "learning_rate": 4.4492576689361705e-06,
+      "loss": 0.1599,
+      "step": 1130
+    },
+    {
+      "epoch": 0.2548229926493368,
+      "grad_norm": 0.5079568505038684,
+      "learning_rate": 4.448090674777377e-06,
+      "loss": 0.1823,
+      "step": 1131
+    },
+    {
+      "epoch": 0.2550483003351452,
+      "grad_norm": 0.49772082318847466,
+      "learning_rate": 4.446922598916445e-06,
+      "loss": 0.1695,
+      "step": 1132
+    },
+    {
+      "epoch": 0.2552736080209536,
+      "grad_norm": 0.4698431427484417,
+      "learning_rate": 4.4457534420019644e-06,
+      "loss": 0.1637,
+      "step": 1133
+    },
+    {
+      "epoch": 0.25549891570676203,
+      "grad_norm": 0.46380041243007897,
+      "learning_rate": 4.444583204683123e-06,
+      "loss": 0.1797,
+      "step": 1134
+    },
+    {
+      "epoch": 0.2557242233925705,
+      "grad_norm": 0.4697980613946506,
+      "learning_rate": 4.44341188760971e-06,
+      "loss": 0.1632,
+      "step": 1135
+    },
+    {
+      "epoch": 0.2559495310783789,
+      "grad_norm": 0.4872747387611847,
+      "learning_rate": 4.4422394914321145e-06,
+      "loss": 0.1778,
+      "step": 1136
+    },
+    {
+      "epoch": 0.25617483876418734,
+      "grad_norm": 0.485343327730141,
+      "learning_rate": 4.4410660168013255e-06,
+      "loss": 0.171,
+      "step": 1137
+    },
+    {
+      "epoch": 0.25640014644999576,
+      "grad_norm": 0.4391103312259274,
+      "learning_rate": 4.439891464368927e-06,
+      "loss": 0.1644,
+      "step": 1138
+    },
+    {
+      "epoch": 0.25662545413580423,
+      "grad_norm": 0.46132359473885676,
+      "learning_rate": 4.438715834787107e-06,
+      "loss": 0.1701,
+      "step": 1139
+    },
+    {
+      "epoch": 0.25685076182161265,
+      "grad_norm": 0.4947273838629531,
+      "learning_rate": 4.437539128708647e-06,
+      "loss": 0.1766,
+      "step": 1140
+    },
+    {
+      "epoch": 0.25707606950742107,
+      "grad_norm": 0.44906983895296854,
+      "learning_rate": 4.436361346786929e-06,
+      "loss": 0.1631,
+      "step": 1141
+    },
+    {
+      "epoch": 0.2573013771932295,
+      "grad_norm": 0.4603708623495546,
+      "learning_rate": 4.435182489675931e-06,
+      "loss": 0.1599,
+      "step": 1142
+    },
+    {
+      "epoch": 0.25752668487903796,
+      "grad_norm": 0.45802671302164594,
+      "learning_rate": 4.4340025580302285e-06,
+      "loss": 0.1729,
+      "step": 1143
+    },
+    {
+      "epoch": 0.2577519925648464,
+      "grad_norm": 0.4650391611868575,
+      "learning_rate": 4.432821552504994e-06,
+      "loss": 0.1609,
+      "step": 1144
+    },
+    {
+      "epoch": 0.2579773002506548,
+      "grad_norm": 0.47776899229826364,
+      "learning_rate": 4.431639473755994e-06,
+      "loss": 0.1692,
+      "step": 1145
+    },
+    {
+      "epoch": 0.2582026079364632,
+      "grad_norm": 0.4748238799376714,
+      "learning_rate": 4.430456322439596e-06,
+      "loss": 0.1711,
+      "step": 1146
+    },
+    {
+      "epoch": 0.2584279156222717,
+      "grad_norm": 0.4965122101302463,
+      "learning_rate": 4.429272099212757e-06,
+      "loss": 0.1768,
+      "step": 1147
+    },
+    {
+      "epoch": 0.2586532233080801,
+      "grad_norm": 0.47030228526971696,
+      "learning_rate": 4.4280868047330325e-06,
+      "loss": 0.163,
+      "step": 1148
+    },
+    {
+      "epoch": 0.2588785309938885,
+      "grad_norm": 0.5011109705302049,
+      "learning_rate": 4.4269004396585735e-06,
+      "loss": 0.1799,
+      "step": 1149
+    },
+    {
+      "epoch": 0.25910383867969694,
+      "grad_norm": 0.4918342025169798,
+      "learning_rate": 4.425713004648123e-06,
+      "loss": 0.1645,
+      "step": 1150
+    },
+    {
+      "epoch": 0.2593291463655054,
+      "grad_norm": 0.4592219591074875,
+      "learning_rate": 4.424524500361021e-06,
+      "loss": 0.1554,
+      "step": 1151
+    },
+    {
+      "epoch": 0.25955445405131383,
+      "grad_norm": 0.4759222622101723,
+      "learning_rate": 4.423334927457198e-06,
+      "loss": 0.1596,
+      "step": 1152
+    },
+    {
+      "epoch": 0.25977976173712225,
+      "grad_norm": 0.49223761810181915,
+      "learning_rate": 4.42214428659718e-06,
+      "loss": 0.1767,
+      "step": 1153
+    },
+    {
+      "epoch": 0.26000506942293067,
+      "grad_norm": 0.4728211158131068,
+      "learning_rate": 4.420952578442086e-06,
+      "loss": 0.1693,
+      "step": 1154
+    },
+    {
+      "epoch": 0.26023037710873914,
+      "grad_norm": 0.4899834125737949,
+      "learning_rate": 4.419759803653627e-06,
+      "loss": 0.1756,
+      "step": 1155
+    },
+    {
+      "epoch": 0.26045568479454756,
+      "grad_norm": 0.5042200638482935,
+      "learning_rate": 4.4185659628941054e-06,
+      "loss": 0.172,
+      "step": 1156
+    },
+    {
+      "epoch": 0.260680992480356,
+      "grad_norm": 0.4381712609704442,
+      "learning_rate": 4.417371056826417e-06,
+      "loss": 0.1512,
+      "step": 1157
+    },
+    {
+      "epoch": 0.2609063001661644,
+      "grad_norm": 0.49604617909817605,
+      "learning_rate": 4.416175086114049e-06,
+      "loss": 0.1821,
+      "step": 1158
+    },
+    {
+      "epoch": 0.26113160785197287,
+      "grad_norm": 0.4805931915635769,
+      "learning_rate": 4.414978051421081e-06,
+      "loss": 0.1751,
+      "step": 1159
+    },
+    {
+      "epoch": 0.2613569155377813,
+      "grad_norm": 0.5108089694144428,
+      "learning_rate": 4.4137799534121785e-06,
+      "loss": 0.178,
+      "step": 1160
+    },
+    {
+      "epoch": 0.2615822232235897,
+      "grad_norm": 0.47149666287470404,
+      "learning_rate": 4.412580792752601e-06,
+      "loss": 0.1698,
+      "step": 1161
+    },
+    {
+      "epoch": 0.2618075309093981,
+      "grad_norm": 0.46335594977891076,
+      "learning_rate": 4.4113805701082e-06,
+      "loss": 0.1702,
+      "step": 1162
+    },
+    {
+      "epoch": 0.2620328385952066,
+      "grad_norm": 0.48331517281476644,
+      "learning_rate": 4.410179286145414e-06,
+      "loss": 0.1643,
+      "step": 1163
+    },
+    {
+      "epoch": 0.262258146281015,
+      "grad_norm": 0.45649384365543133,
+      "learning_rate": 4.408976941531269e-06,
+      "loss": 0.1731,
+      "step": 1164
+    },
+    {
+      "epoch": 0.26248345396682343,
+      "grad_norm": 0.48853109585960214,
+      "learning_rate": 4.407773536933384e-06,
+      "loss": 0.1872,
+      "step": 1165
+    },
+    {
+      "epoch": 0.26270876165263185,
+      "grad_norm": 0.47250454754886273,
+      "learning_rate": 4.406569073019965e-06,
+      "loss": 0.1739,
+      "step": 1166
+    },
+    {
+      "epoch": 0.2629340693384403,
+      "grad_norm": 0.4868051127801619,
+      "learning_rate": 4.4053635504598045e-06,
+      "loss": 0.1725,
+      "step": 1167
+    },
+    {
+      "epoch": 0.26315937702424874,
+      "grad_norm": 0.46804005017622746,
+      "learning_rate": 4.404156969922284e-06,
+      "loss": 0.1702,
+      "step": 1168
+    },
+    {
+      "epoch": 0.26338468471005716,
+      "grad_norm": 0.4953399325232565,
+      "learning_rate": 4.402949332077375e-06,
+      "loss": 0.1727,
+      "step": 1169
+    },
+    {
+      "epoch": 0.2636099923958656,
+      "grad_norm": 0.5050035534953599,
+      "learning_rate": 4.401740637595633e-06,
+      "loss": 0.1638,
+      "step": 1170
+    },
+    {
+      "epoch": 0.26383530008167405,
+      "grad_norm": 0.5061533696389707,
+      "learning_rate": 4.400530887148199e-06,
+      "loss": 0.1678,
+      "step": 1171
+    },
+    {
+      "epoch": 0.26406060776748247,
+      "grad_norm": 0.48214334053632324,
+      "learning_rate": 4.3993200814068035e-06,
+      "loss": 0.1592,
+      "step": 1172
+    },
+    {
+      "epoch": 0.2642859154532909,
+      "grad_norm": 0.47051364612550667,
+      "learning_rate": 4.398108221043764e-06,
+      "loss": 0.1843,
+      "step": 1173
+    },
+    {
+      "epoch": 0.2645112231390993,
+      "grad_norm": 0.47107601682684574,
+      "learning_rate": 4.396895306731978e-06,
+      "loss": 0.173,
+      "step": 1174
+    },
+    {
+      "epoch": 0.2647365308249078,
+      "grad_norm": 0.5005898656550045,
+      "learning_rate": 4.395681339144933e-06,
+      "loss": 0.1902,
+      "step": 1175
+    },
+    {
+      "epoch": 0.2649618385107162,
+      "grad_norm": 0.4676621595113707,
+      "learning_rate": 4.394466318956701e-06,
+      "loss": 0.1656,
+      "step": 1176
+    },
+    {
+      "epoch": 0.2651871461965246,
+      "grad_norm": 0.49204892601984884,
+      "learning_rate": 4.393250246841935e-06,
+      "loss": 0.1693,
+      "step": 1177
+    },
+    {
+      "epoch": 0.2654124538823331,
+      "grad_norm": 0.46310285597788936,
+      "learning_rate": 4.392033123475876e-06,
+      "loss": 0.1681,
+      "step": 1178
+    },
+    {
+      "epoch": 0.2656377615681415,
+      "grad_norm": 0.47878735086717533,
+      "learning_rate": 4.390814949534348e-06,
+      "loss": 0.168,
+      "step": 1179
+    },
+    {
+      "epoch": 0.2658630692539499,
+      "grad_norm": 0.49825319716859656,
+      "learning_rate": 4.389595725693756e-06,
+      "loss": 0.1696,
+      "step": 1180
+    },
+    {
+      "epoch": 0.26608837693975834,
+      "grad_norm": 0.4455782342121541,
+      "learning_rate": 4.388375452631091e-06,
+      "loss": 0.1615,
+      "step": 1181
+    },
+    {
+      "epoch": 0.2663136846255668,
+      "grad_norm": 0.43357859168921514,
+      "learning_rate": 4.387154131023924e-06,
+      "loss": 0.1551,
+      "step": 1182
+    },
+    {
+      "epoch": 0.26653899231137523,
+      "grad_norm": 0.5089345638479391,
+      "learning_rate": 4.385931761550411e-06,
+      "loss": 0.176,
+      "step": 1183
+    },
+    {
+      "epoch": 0.26676429999718365,
+      "grad_norm": 0.5266595479693508,
+      "learning_rate": 4.384708344889285e-06,
+      "loss": 0.1695,
+      "step": 1184
+    },
+    {
+      "epoch": 0.26698960768299207,
+      "grad_norm": 0.475099637405834,
+      "learning_rate": 4.383483881719867e-06,
+      "loss": 0.1779,
+      "step": 1185
+    },
+    {
+      "epoch": 0.26721491536880054,
+      "grad_norm": 0.4787739839641768,
+      "learning_rate": 4.382258372722054e-06,
+      "loss": 0.1725,
+      "step": 1186
+    },
+    {
+      "epoch": 0.26744022305460896,
+      "grad_norm": 0.49494278847012774,
+      "learning_rate": 4.381031818576326e-06,
+      "loss": 0.165,
+      "step": 1187
+    },
+    {
+      "epoch": 0.2676655307404174,
+      "grad_norm": 0.46344787955776795,
+      "learning_rate": 4.379804219963742e-06,
+      "loss": 0.1673,
+      "step": 1188
+    },
+    {
+      "epoch": 0.2678908384262258,
+      "grad_norm": 0.46819728716710135,
+      "learning_rate": 4.378575577565945e-06,
+      "loss": 0.1685,
+      "step": 1189
+    },
+    {
+      "epoch": 0.26811614611203427,
+      "grad_norm": 0.4634414488798126,
+      "learning_rate": 4.377345892065149e-06,
+      "loss": 0.1572,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2683414537978427,
+      "grad_norm": 0.47565010962702786,
+      "learning_rate": 4.376115164144157e-06,
+      "loss": 0.1718,
+      "step": 1191
+    },
+    {
+      "epoch": 0.2685667614836511,
+      "grad_norm": 0.47429751678202337,
+      "learning_rate": 4.374883394486343e-06,
+      "loss": 0.1706,
+      "step": 1192
+    },
+    {
+      "epoch": 0.2687920691694595,
+      "grad_norm": 0.5085699899597834,
+      "learning_rate": 4.373650583775666e-06,
+      "loss": 0.1648,
+      "step": 1193
+    },
+    {
+      "epoch": 0.269017376855268,
+      "grad_norm": 0.5146572632392027,
+      "learning_rate": 4.3724167326966575e-06,
+      "loss": 0.1878,
+      "step": 1194
+    },
+    {
+      "epoch": 0.2692426845410764,
+      "grad_norm": 0.4627246459636679,
+      "learning_rate": 4.37118184193443e-06,
+      "loss": 0.1572,
+      "step": 1195
+    },
+    {
+      "epoch": 0.26946799222688483,
+      "grad_norm": 0.5042679214482405,
+      "learning_rate": 4.3699459121746726e-06,
+      "loss": 0.1789,
+      "step": 1196
+    },
+    {
+      "epoch": 0.26969329991269325,
+      "grad_norm": 0.4715362603580668,
+      "learning_rate": 4.368708944103649e-06,
+      "loss": 0.1565,
+      "step": 1197
+    },
+    {
+      "epoch": 0.2699186075985017,
+      "grad_norm": 0.46224782361941275,
+      "learning_rate": 4.367470938408204e-06,
+      "loss": 0.1638,
+      "step": 1198
+    },
+    {
+      "epoch": 0.27014391528431014,
+      "grad_norm": 0.48457161847040103,
+      "learning_rate": 4.366231895775755e-06,
+      "loss": 0.1697,
+      "step": 1199
+    },
+    {
+      "epoch": 0.27036922297011856,
+      "grad_norm": 0.45322439544830784,
+      "learning_rate": 4.364991816894296e-06,
+      "loss": 0.1492,
+      "step": 1200
+    },
+    {
+      "epoch": 0.270594530655927,
+      "grad_norm": 0.5475044658907094,
+      "learning_rate": 4.3637507024523975e-06,
+      "loss": 0.1988,
+      "step": 1201
+    },
+    {
+      "epoch": 0.27081983834173545,
+      "grad_norm": 0.4898131791403852,
+      "learning_rate": 4.362508553139203e-06,
+      "loss": 0.1777,
+      "step": 1202
+    },
+    {
+      "epoch": 0.27104514602754387,
+      "grad_norm": 0.4540719396050229,
+      "learning_rate": 4.361265369644432e-06,
+      "loss": 0.1652,
+      "step": 1203
+    },
+    {
+      "epoch": 0.2712704537133523,
+      "grad_norm": 0.4827482306962776,
+      "learning_rate": 4.360021152658378e-06,
+      "loss": 0.1797,
+      "step": 1204
+    },
+    {
+      "epoch": 0.2714957613991607,
+      "grad_norm": 0.465499233492623,
+      "learning_rate": 4.3587759028719075e-06,
+      "loss": 0.1607,
+      "step": 1205
+    },
+    {
+      "epoch": 0.2717210690849692,
+      "grad_norm": 0.45932842008835173,
+      "learning_rate": 4.357529620976463e-06,
+      "loss": 0.1658,
+      "step": 1206
+    },
+    {
+      "epoch": 0.2719463767707776,
+      "grad_norm": 0.49105618480613644,
+      "learning_rate": 4.356282307664057e-06,
+      "loss": 0.1708,
+      "step": 1207
+    },
+    {
+      "epoch": 0.272171684456586,
+      "grad_norm": 0.4699501932404702,
+      "learning_rate": 4.355033963627277e-06,
+      "loss": 0.1623,
+      "step": 1208
+    },
+    {
+      "epoch": 0.27239699214239443,
+      "grad_norm": 0.4432815322488692,
+      "learning_rate": 4.353784589559282e-06,
+      "loss": 0.1523,
+      "step": 1209
+    },
+    {
+      "epoch": 0.2726222998282029,
+      "grad_norm": 0.4999895546959578,
+      "learning_rate": 4.352534186153802e-06,
+      "loss": 0.1714,
+      "step": 1210
+    },
+    {
+      "epoch": 0.2728476075140113,
+      "grad_norm": 0.5376208111714382,
+      "learning_rate": 4.35128275410514e-06,
+      "loss": 0.1684,
+      "step": 1211
+    },
+    {
+      "epoch": 0.27307291519981974,
+      "grad_norm": 0.5005445898532112,
+      "learning_rate": 4.3500302941081685e-06,
+      "loss": 0.1766,
+      "step": 1212
+    },
+    {
+      "epoch": 0.27329822288562816,
+      "grad_norm": 0.4916858385322702,
+      "learning_rate": 4.348776806858334e-06,
+      "loss": 0.1746,
+      "step": 1213
+    },
+    {
+      "epoch": 0.27352353057143663,
+      "grad_norm": 0.5038191015625039,
+      "learning_rate": 4.3475222930516484e-06,
+      "loss": 0.1761,
+      "step": 1214
+    },
+    {
+      "epoch": 0.27374883825724505,
+      "grad_norm": 0.482264627560053,
+      "learning_rate": 4.346266753384699e-06,
+      "loss": 0.166,
+      "step": 1215
+    },
+    {
+      "epoch": 0.27397414594305347,
+      "grad_norm": 0.4782417492804113,
+      "learning_rate": 4.345010188554638e-06,
+      "loss": 0.1682,
+      "step": 1216
+    },
+    {
+      "epoch": 0.2741994536288619,
+      "grad_norm": 0.5225392937950303,
+      "learning_rate": 4.343752599259192e-06,
+      "loss": 0.1775,
+      "step": 1217
+    },
+    {
+      "epoch": 0.27442476131467036,
+      "grad_norm": 0.46930424506116875,
+      "learning_rate": 4.34249398619665e-06,
+      "loss": 0.1685,
+      "step": 1218
+    },
+    {
+      "epoch": 0.2746500690004788,
+      "grad_norm": 0.4810395604573289,
+      "learning_rate": 4.341234350065876e-06,
+      "loss": 0.1804,
+      "step": 1219
+    },
+    {
+      "epoch": 0.2748753766862872,
+      "grad_norm": 0.457383636891631,
+      "learning_rate": 4.339973691566297e-06,
+      "loss": 0.1616,
+      "step": 1220
+    },
+    {
+      "epoch": 0.2751006843720956,
+      "grad_norm": 0.4711529416779021,
+      "learning_rate": 4.33871201139791e-06,
+      "loss": 0.1617,
+      "step": 1221
+    },
+    {
+      "epoch": 0.2753259920579041,
+      "grad_norm": 0.5160260790743614,
+      "learning_rate": 4.337449310261279e-06,
+      "loss": 0.1684,
+      "step": 1222
+    },
+    {
+      "epoch": 0.2755512997437125,
+      "grad_norm": 0.517349396391902,
+      "learning_rate": 4.336185588857535e-06,
+      "loss": 0.1765,
+      "step": 1223
+    },
+    {
+      "epoch": 0.2757766074295209,
+      "grad_norm": 0.49587416173024024,
+      "learning_rate": 4.334920847888376e-06,
+      "loss": 0.1709,
+      "step": 1224
+    },
+    {
+      "epoch": 0.2760019151153294,
+      "grad_norm": 0.48410326600056597,
+      "learning_rate": 4.333655088056065e-06,
+      "loss": 0.1767,
+      "step": 1225
+    },
+    {
+      "epoch": 0.2762272228011378,
+      "grad_norm": 0.5468232901412815,
+      "learning_rate": 4.332388310063431e-06,
+      "loss": 0.1876,
+      "step": 1226
+    },
+    {
+      "epoch": 0.27645253048694624,
+      "grad_norm": 0.5317933634678655,
+      "learning_rate": 4.331120514613869e-06,
+      "loss": 0.1727,
+      "step": 1227
+    },
+    {
+      "epoch": 0.27667783817275465,
+      "grad_norm": 0.5022448146956019,
+      "learning_rate": 4.329851702411339e-06,
+      "loss": 0.1721,
+      "step": 1228
+    },
+    {
+      "epoch": 0.2769031458585631,
+      "grad_norm": 0.46481136517383537,
+      "learning_rate": 4.328581874160363e-06,
+      "loss": 0.1558,
+      "step": 1229
+    },
+    {
+      "epoch": 0.27712845354437154,
+      "grad_norm": 0.5284031800079505,
+      "learning_rate": 4.327311030566033e-06,
+      "loss": 0.1578,
+      "step": 1230
+    },
+    {
+      "epoch": 0.27735376123017996,
+      "grad_norm": 0.5138307368684882,
+      "learning_rate": 4.326039172333997e-06,
+      "loss": 0.1783,
+      "step": 1231
+    },
+    {
+      "epoch": 0.2775790689159884,
+      "grad_norm": 0.5042489855251965,
+      "learning_rate": 4.324766300170473e-06,
+      "loss": 0.1771,
+      "step": 1232
+    },
+    {
+      "epoch": 0.27780437660179685,
+      "grad_norm": 0.5161807185989821,
+      "learning_rate": 4.323492414782239e-06,
+      "loss": 0.1695,
+      "step": 1233
+    },
+    {
+      "epoch": 0.27802968428760527,
+      "grad_norm": 0.49481011856084106,
+      "learning_rate": 4.322217516876635e-06,
+      "loss": 0.1584,
+      "step": 1234
+    },
+    {
+      "epoch": 0.2782549919734137,
+      "grad_norm": 0.5297983704670794,
+      "learning_rate": 4.320941607161567e-06,
+      "loss": 0.1819,
+      "step": 1235
+    },
+    {
+      "epoch": 0.2784802996592221,
+      "grad_norm": 0.504028057895518,
+      "learning_rate": 4.3196646863454975e-06,
+      "loss": 0.1692,
+      "step": 1236
+    },
+    {
+      "epoch": 0.2787056073450306,
+      "grad_norm": 0.48754978064018856,
+      "learning_rate": 4.3183867551374535e-06,
+      "loss": 0.1696,
+      "step": 1237
+    },
+    {
+      "epoch": 0.278930915030839,
+      "grad_norm": 0.5186780873159246,
+      "learning_rate": 4.317107814247022e-06,
+      "loss": 0.1887,
+      "step": 1238
+    },
+    {
+      "epoch": 0.2791562227166474,
+      "grad_norm": 0.49208775545274297,
+      "learning_rate": 4.3158278643843544e-06,
+      "loss": 0.1772,
+      "step": 1239
+    },
+    {
+      "epoch": 0.27938153040245584,
+      "grad_norm": 0.4834696449410449,
+      "learning_rate": 4.314546906260156e-06,
+      "loss": 0.1743,
+      "step": 1240
+    },
+    {
+      "epoch": 0.2796068380882643,
+      "grad_norm": 0.4940735421374972,
+      "learning_rate": 4.313264940585695e-06,
+      "loss": 0.1657,
+      "step": 1241
+    },
+    {
+      "epoch": 0.2798321457740727,
+      "grad_norm": 0.48222847794101337,
+      "learning_rate": 4.3119819680728e-06,
+      "loss": 0.1571,
+      "step": 1242
+    },
+    {
+      "epoch": 0.28005745345988114,
+      "grad_norm": 0.504598014321156,
+      "learning_rate": 4.310697989433858e-06,
+      "loss": 0.1766,
+      "step": 1243
+    },
+    {
+      "epoch": 0.28028276114568956,
+      "grad_norm": 0.4661881543796616,
+      "learning_rate": 4.3094130053818164e-06,
+      "loss": 0.1629,
+      "step": 1244
+    },
+    {
+      "epoch": 0.28050806883149804,
+      "grad_norm": 0.48427367373674574,
+      "learning_rate": 4.308127016630176e-06,
+      "loss": 0.1695,
+      "step": 1245
+    },
+    {
+      "epoch": 0.28073337651730645,
+      "grad_norm": 0.5031029441970696,
+      "learning_rate": 4.306840023892998e-06,
+      "loss": 0.1697,
+      "step": 1246
+    },
+    {
+      "epoch": 0.2809586842031149,
+      "grad_norm": 0.4682614760551946,
+      "learning_rate": 4.305552027884904e-06,
+      "loss": 0.1615,
+      "step": 1247
+    },
+    {
+      "epoch": 0.2811839918889233,
+      "grad_norm": 0.46375682682480657,
+      "learning_rate": 4.304263029321069e-06,
+      "loss": 0.1721,
+      "step": 1248
+    },
+    {
+      "epoch": 0.28140929957473176,
+      "grad_norm": 0.46393825922775633,
+      "learning_rate": 4.302973028917226e-06,
+      "loss": 0.1762,
+      "step": 1249
+    },
+    {
+      "epoch": 0.2816346072605402,
+      "grad_norm": 0.46920503962002086,
+      "learning_rate": 4.301682027389663e-06,
+      "loss": 0.1721,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2818599149463486,
+      "grad_norm": 0.48383966890417285,
+      "learning_rate": 4.300390025455227e-06,
+      "loss": 0.1679,
+      "step": 1251
+    },
+    {
+      "epoch": 0.282085222632157,
+      "grad_norm": 0.4950523538046429,
+      "learning_rate": 4.299097023831318e-06,
+      "loss": 0.1671,
+      "step": 1252
+    },
+    {
+      "epoch": 0.2823105303179655,
+      "grad_norm": 0.45179000541068043,
+      "learning_rate": 4.2978030232358904e-06,
+      "loss": 0.1684,
+      "step": 1253
+    },
+    {
+      "epoch": 0.2825358380037739,
+      "grad_norm": 0.4569613433988033,
+      "learning_rate": 4.2965080243874555e-06,
+      "loss": 0.1561,
+      "step": 1254
+    },
+    {
+      "epoch": 0.2827611456895823,
+      "grad_norm": 0.49036731780666165,
+      "learning_rate": 4.295212028005078e-06,
+      "loss": 0.1771,
+      "step": 1255
+    },
+    {
+      "epoch": 0.28298645337539075,
+      "grad_norm": 0.4504370205410457,
+      "learning_rate": 4.293915034808376e-06,
+      "loss": 0.1557,
+      "step": 1256
+    },
+    {
+      "epoch": 0.2832117610611992,
+      "grad_norm": 0.47106233823900273,
+      "learning_rate": 4.292617045517521e-06,
+      "loss": 0.1606,
+      "step": 1257
+    },
+    {
+      "epoch": 0.28343706874700764,
+      "grad_norm": 0.4313797520186185,
+      "learning_rate": 4.29131806085324e-06,
+      "loss": 0.1452,
+      "step": 1258
+    },
+    {
+      "epoch": 0.28366237643281605,
+      "grad_norm": 0.4568166083233466,
+      "learning_rate": 4.290018081536807e-06,
+      "loss": 0.1629,
+      "step": 1259
+    },
+    {
+      "epoch": 0.2838876841186245,
+      "grad_norm": 0.47667485262652165,
+      "learning_rate": 4.288717108290056e-06,
+      "loss": 0.1626,
+      "step": 1260
+    },
+    {
+      "epoch": 0.28411299180443295,
+      "grad_norm": 0.4807412272846728,
+      "learning_rate": 4.287415141835368e-06,
+      "loss": 0.1846,
+      "step": 1261
+    },
+    {
+      "epoch": 0.28433829949024136,
+      "grad_norm": 0.4681601937744111,
+      "learning_rate": 4.2861121828956745e-06,
+      "loss": 0.1701,
+      "step": 1262
+    },
+    {
+      "epoch": 0.2845636071760498,
+      "grad_norm": 0.48879201661900573,
+      "learning_rate": 4.284808232194462e-06,
+      "loss": 0.1765,
+      "step": 1263
+    },
+    {
+      "epoch": 0.2847889148618582,
+      "grad_norm": 0.4676260132902741,
+      "learning_rate": 4.283503290455765e-06,
+      "loss": 0.1676,
+      "step": 1264
+    },
+    {
+      "epoch": 0.2850142225476667,
+      "grad_norm": 0.49374539412848295,
+      "learning_rate": 4.28219735840417e-06,
+      "loss": 0.1711,
+      "step": 1265
+    },
+    {
+      "epoch": 0.2852395302334751,
+      "grad_norm": 0.5133291385107003,
+      "learning_rate": 4.28089043676481e-06,
+      "loss": 0.1751,
+      "step": 1266
+    },
+    {
+      "epoch": 0.2854648379192835,
+      "grad_norm": 0.5015169344912714,
+      "learning_rate": 4.279582526263371e-06,
+      "loss": 0.1799,
+      "step": 1267
+    },
+    {
+      "epoch": 0.2856901456050919,
+      "grad_norm": 0.49564604641852966,
+      "learning_rate": 4.27827362762609e-06,
+      "loss": 0.1868,
+      "step": 1268
+    },
+    {
+      "epoch": 0.2859154532909004,
+      "grad_norm": 0.4826518697709936,
+      "learning_rate": 4.276963741579745e-06,
+      "loss": 0.1724,
+      "step": 1269
+    },
+    {
+      "epoch": 0.2861407609767088,
+      "grad_norm": 0.47198024006927375,
+      "learning_rate": 4.275652868851669e-06,
+      "loss": 0.1626,
+      "step": 1270
+    },
+    {
+      "epoch": 0.28636606866251724,
+      "grad_norm": 0.4887317789419866,
+      "learning_rate": 4.2743410101697405e-06,
+      "loss": 0.1759,
+      "step": 1271
+    },
+    {
+      "epoch": 0.28659137634832565,
+      "grad_norm": 0.48415900082739516,
+      "learning_rate": 4.2730281662623866e-06,
+      "loss": 0.1731,
+      "step": 1272
+    },
+    {
+      "epoch": 0.28681668403413413,
+      "grad_norm": 0.48162186592846623,
+      "learning_rate": 4.271714337858579e-06,
+      "loss": 0.1681,
+      "step": 1273
+    },
+    {
+      "epoch": 0.28704199171994255,
+      "grad_norm": 0.4555532662421832,
+      "learning_rate": 4.270399525687839e-06,
+      "loss": 0.1653,
+      "step": 1274
+    },
+    {
+      "epoch": 0.28726729940575096,
+      "grad_norm": 0.49380043518937744,
+      "learning_rate": 4.269083730480232e-06,
+      "loss": 0.1841,
+      "step": 1275
+    },
+    {
+      "epoch": 0.28749260709155944,
+      "grad_norm": 0.470663539027122,
+      "learning_rate": 4.267766952966369e-06,
+      "loss": 0.1682,
+      "step": 1276
+    },
+    {
+      "epoch": 0.28771791477736786,
+      "grad_norm": 0.4861182617285624,
+      "learning_rate": 4.26644919387741e-06,
+      "loss": 0.1688,
+      "step": 1277
+    },
+    {
+      "epoch": 0.2879432224631763,
+      "grad_norm": 0.45164008956634005,
+      "learning_rate": 4.265130453945056e-06,
+      "loss": 0.1592,
+      "step": 1278
+    },
+    {
+      "epoch": 0.2881685301489847,
+      "grad_norm": 0.4813784687003862,
+      "learning_rate": 4.263810733901554e-06,
+      "loss": 0.1819,
+      "step": 1279
+    },
+    {
+      "epoch": 0.28839383783479317,
+      "grad_norm": 0.4737265609854097,
+      "learning_rate": 4.262490034479697e-06,
+      "loss": 0.1662,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2886191455206016,
+      "grad_norm": 0.44082202687660105,
+      "learning_rate": 4.261168356412818e-06,
+      "loss": 0.1631,
+      "step": 1281
+    },
+    {
+      "epoch": 0.28884445320641,
+      "grad_norm": 0.43808182321709044,
+      "learning_rate": 4.259845700434797e-06,
+      "loss": 0.1473,
+      "step": 1282
+    },
+    {
+      "epoch": 0.2890697608922184,
+      "grad_norm": 0.5104052734846787,
+      "learning_rate": 4.258522067280055e-06,
+      "loss": 0.171,
+      "step": 1283
+    },
+    {
+      "epoch": 0.2892950685780269,
+      "grad_norm": 0.48126358259683144,
+      "learning_rate": 4.257197457683556e-06,
+      "loss": 0.1602,
+      "step": 1284
+    },
+    {
+      "epoch": 0.2895203762638353,
+      "grad_norm": 0.4860327782364687,
+      "learning_rate": 4.2558718723808055e-06,
+      "loss": 0.1777,
+      "step": 1285
+    },
+    {
+      "epoch": 0.28974568394964373,
+      "grad_norm": 0.4909158607880032,
+      "learning_rate": 4.254545312107854e-06,
+      "loss": 0.1749,
+      "step": 1286
+    },
+    {
+      "epoch": 0.28997099163545215,
+      "grad_norm": 0.515698974564528,
+      "learning_rate": 4.253217777601289e-06,
+      "loss": 0.1832,
+      "step": 1287
+    },
+    {
+      "epoch": 0.2901962993212606,
+      "grad_norm": 0.49937903200840367,
+      "learning_rate": 4.251889269598241e-06,
+      "loss": 0.1877,
+      "step": 1288
+    },
+    {
+      "epoch": 0.29042160700706904,
+      "grad_norm": 0.4583839347106229,
+      "learning_rate": 4.250559788836382e-06,
+      "loss": 0.1636,
+      "step": 1289
+    },
+    {
+      "epoch": 0.29064691469287746,
+      "grad_norm": 0.4891595487665644,
+      "learning_rate": 4.249229336053924e-06,
+      "loss": 0.1742,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2908722223786859,
+      "grad_norm": 0.46454308268647515,
+      "learning_rate": 4.247897911989615e-06,
+      "loss": 0.1672,
+      "step": 1291
+    },
+    {
+      "epoch": 0.29109753006449435,
+      "grad_norm": 0.45860463509021276,
+      "learning_rate": 4.2465655173827465e-06,
+      "loss": 0.1687,
+      "step": 1292
+    },
+    {
+      "epoch": 0.29132283775030277,
+      "grad_norm": 0.46340134333777605,
+      "learning_rate": 4.245232152973148e-06,
+      "loss": 0.1575,
+      "step": 1293
+    },
+    {
+      "epoch": 0.2915481454361112,
+      "grad_norm": 0.4709196024735463,
+      "learning_rate": 4.243897819501187e-06,
+      "loss": 0.156,
+      "step": 1294
+    },
+    {
+      "epoch": 0.2917734531219196,
+      "grad_norm": 0.4720281820205427,
+      "learning_rate": 4.242562517707768e-06,
+      "loss": 0.1607,
+      "step": 1295
+    },
+    {
+      "epoch": 0.2919987608077281,
+      "grad_norm": 0.4950495777319756,
+      "learning_rate": 4.241226248334335e-06,
+      "loss": 0.1771,
+      "step": 1296
+    },
+    {
+      "epoch": 0.2922240684935365,
+      "grad_norm": 0.44321757258749245,
+      "learning_rate": 4.23988901212287e-06,
+      "loss": 0.1501,
+      "step": 1297
+    },
+    {
+      "epoch": 0.2924493761793449,
+      "grad_norm": 0.46641258630583354,
+      "learning_rate": 4.238550809815889e-06,
+      "loss": 0.1577,
+      "step": 1298
+    },
+    {
+      "epoch": 0.29267468386515333,
+      "grad_norm": 0.4355684264835874,
+      "learning_rate": 4.237211642156446e-06,
+      "loss": 0.1539,
+      "step": 1299
+    },
+    {
+      "epoch": 0.2928999915509618,
+      "grad_norm": 0.46022050786059787,
+      "learning_rate": 4.23587150988813e-06,
+      "loss": 0.1657,
+      "step": 1300
+    },
+    {
+      "epoch": 0.2931252992367702,
+      "grad_norm": 0.50555996380803,
+      "learning_rate": 4.234530413755069e-06,
+      "loss": 0.1861,
+      "step": 1301
+    },
+    {
+      "epoch": 0.29335060692257864,
+      "grad_norm": 0.45770806516472096,
+      "learning_rate": 4.233188354501921e-06,
+      "loss": 0.1658,
+      "step": 1302
+    },
+    {
+      "epoch": 0.29357591460838706,
+      "grad_norm": 0.4822249847330653,
+      "learning_rate": 4.231845332873883e-06,
+      "loss": 0.1675,
+      "step": 1303
+    },
+    {
+      "epoch": 0.29380122229419553,
+      "grad_norm": 0.47560738464013713,
+      "learning_rate": 4.230501349616683e-06,
+      "loss": 0.1639,
+      "step": 1304
+    },
+    {
+      "epoch": 0.29402652998000395,
+      "grad_norm": 0.486640763683581,
+      "learning_rate": 4.2291564054765876e-06,
+      "loss": 0.1708,
+      "step": 1305
+    },
+    {
+      "epoch": 0.29425183766581237,
+      "grad_norm": 0.4772344480015269,
+      "learning_rate": 4.227810501200393e-06,
+      "loss": 0.1705,
+      "step": 1306
+    },
+    {
+      "epoch": 0.2944771453516208,
+      "grad_norm": 0.49761222589037885,
+      "learning_rate": 4.226463637535429e-06,
+      "loss": 0.1856,
+      "step": 1307
+    },
+    {
+      "epoch": 0.29470245303742926,
+      "grad_norm": 0.48453435723802374,
+      "learning_rate": 4.225115815229559e-06,
+      "loss": 0.1709,
+      "step": 1308
+    },
+    {
+      "epoch": 0.2949277607232377,
+      "grad_norm": 0.46816646410647905,
+      "learning_rate": 4.22376703503118e-06,
+      "loss": 0.1676,
+      "step": 1309
+    },
+    {
+      "epoch": 0.2951530684090461,
+      "grad_norm": 0.44980063168003603,
+      "learning_rate": 4.222417297689217e-06,
+      "loss": 0.1492,
+      "step": 1310
+    },
+    {
+      "epoch": 0.2953783760948545,
+      "grad_norm": 0.475968174132624,
+      "learning_rate": 4.22106660395313e-06,
+      "loss": 0.174,
+      "step": 1311
+    },
+    {
+      "epoch": 0.295603683780663,
+      "grad_norm": 0.4966748118697132,
+      "learning_rate": 4.219714954572909e-06,
+      "loss": 0.1704,
+      "step": 1312
+    },
+    {
+      "epoch": 0.2958289914664714,
+      "grad_norm": 0.4864155888357422,
+      "learning_rate": 4.218362350299075e-06,
+      "loss": 0.1754,
+      "step": 1313
+    },
+    {
+      "epoch": 0.2960542991522798,
+      "grad_norm": 0.5153435414046714,
+      "learning_rate": 4.217008791882678e-06,
+      "loss": 0.186,
+      "step": 1314
+    },
+    {
+      "epoch": 0.29627960683808824,
+      "grad_norm": 0.4542256078427076,
+      "learning_rate": 4.215654280075297e-06,
+      "loss": 0.1723,
+      "step": 1315
+    },
+    {
+      "epoch": 0.2965049145238967,
+      "grad_norm": 0.4805611576250163,
+      "learning_rate": 4.214298815629046e-06,
+      "loss": 0.1767,
+      "step": 1316
+    },
+    {
+      "epoch": 0.29673022220970513,
+      "grad_norm": 0.47706024349233667,
+      "learning_rate": 4.212942399296559e-06,
+      "loss": 0.1555,
+      "step": 1317
+    },
+    {
+      "epoch": 0.29695552989551355,
+      "grad_norm": 0.48748152898974967,
+      "learning_rate": 4.211585031831007e-06,
+      "loss": 0.1846,
+      "step": 1318
+    },
+    {
+      "epoch": 0.29718083758132197,
+      "grad_norm": 0.46300199198371683,
+      "learning_rate": 4.210226713986085e-06,
+      "loss": 0.168,
+      "step": 1319
+    },
+    {
+      "epoch": 0.29740614526713044,
+      "grad_norm": 0.4687496463200866,
+      "learning_rate": 4.208867446516015e-06,
+      "loss": 0.1666,
+      "step": 1320
+    },
+    {
+      "epoch": 0.29763145295293886,
+      "grad_norm": 0.46195114020388633,
+      "learning_rate": 4.2075072301755486e-06,
+      "loss": 0.1525,
+      "step": 1321
+    },
+    {
+      "epoch": 0.2978567606387473,
+      "grad_norm": 0.4655949755044446,
+      "learning_rate": 4.206146065719963e-06,
+      "loss": 0.165,
+      "step": 1322
+    },
+    {
+      "epoch": 0.29808206832455575,
+      "grad_norm": 0.476253279421997,
+      "learning_rate": 4.204783953905062e-06,
+      "loss": 0.1787,
+      "step": 1323
+    },
+    {
+      "epoch": 0.29830737601036417,
+      "grad_norm": 0.4483571352663244,
+      "learning_rate": 4.203420895487175e-06,
+      "loss": 0.1661,
+      "step": 1324
+    },
+    {
+      "epoch": 0.2985326836961726,
+      "grad_norm": 0.47256546371837377,
+      "learning_rate": 4.202056891223159e-06,
+      "loss": 0.1659,
+      "step": 1325
+    },
+    {
+      "epoch": 0.298757991381981,
+      "grad_norm": 0.4891184537367563,
+      "learning_rate": 4.200691941870392e-06,
+      "loss": 0.1632,
+      "step": 1326
+    },
+    {
+      "epoch": 0.2989832990677895,
+      "grad_norm": 0.44451431166176403,
+      "learning_rate": 4.199326048186783e-06,
+      "loss": 0.1548,
+      "step": 1327
+    },
+    {
+      "epoch": 0.2992086067535979,
+      "grad_norm": 0.4822388959051435,
+      "learning_rate": 4.197959210930759e-06,
+      "loss": 0.1783,
+      "step": 1328
+    },
+    {
+      "epoch": 0.2994339144394063,
+      "grad_norm": 0.4859493329890237,
+      "learning_rate": 4.196591430861275e-06,
+      "loss": 0.1687,
+      "step": 1329
+    },
+    {
+      "epoch": 0.29965922212521473,
+      "grad_norm": 0.45993151866349613,
+      "learning_rate": 4.195222708737809e-06,
+      "loss": 0.1656,
+      "step": 1330
+    },
+    {
+      "epoch": 0.2998845298110232,
+      "grad_norm": 0.5027414331804484,
+      "learning_rate": 4.193853045320359e-06,
+      "loss": 0.1896,
+      "step": 1331
+    },
+    {
+      "epoch": 0.3001098374968316,
+      "grad_norm": 0.4741251417191081,
+      "learning_rate": 4.192482441369451e-06,
+      "loss": 0.1576,
+      "step": 1332
+    },
+    {
+      "epoch": 0.30033514518264004,
+      "grad_norm": 0.47610034599935547,
+      "learning_rate": 4.191110897646129e-06,
+      "loss": 0.1618,
+      "step": 1333
+    },
+    {
+      "epoch": 0.30056045286844846,
+      "grad_norm": 0.4884119475840316,
+      "learning_rate": 4.189738414911959e-06,
+      "loss": 0.1707,
+      "step": 1334
+    },
+    {
+      "epoch": 0.30078576055425693,
+      "grad_norm": 0.487269387089356,
+      "learning_rate": 4.188364993929032e-06,
+      "loss": 0.1846,
+      "step": 1335
+    },
+    {
+      "epoch": 0.30101106824006535,
+      "grad_norm": 0.4695057148911935,
+      "learning_rate": 4.186990635459954e-06,
+      "loss": 0.1675,
+      "step": 1336
+    },
+    {
+      "epoch": 0.30123637592587377,
+      "grad_norm": 0.4681903788160959,
+      "learning_rate": 4.185615340267858e-06,
+      "loss": 0.1578,
+      "step": 1337
+    },
+    {
+      "epoch": 0.3014616836116822,
+      "grad_norm": 0.4872111988165553,
+      "learning_rate": 4.184239109116393e-06,
+      "loss": 0.1809,
+      "step": 1338
+    },
+    {
+      "epoch": 0.30168699129749066,
+      "grad_norm": 0.44572910840509217,
+      "learning_rate": 4.182861942769729e-06,
+      "loss": 0.1624,
+      "step": 1339
+    },
+    {
+      "epoch": 0.3019122989832991,
+      "grad_norm": 0.46711380169986927,
+      "learning_rate": 4.181483841992556e-06,
+      "loss": 0.1608,
+      "step": 1340
+    },
+    {
+      "epoch": 0.3021376066691075,
+      "grad_norm": 0.4704023219402658,
+      "learning_rate": 4.18010480755008e-06,
+      "loss": 0.1654,
+      "step": 1341
+    },
+    {
+      "epoch": 0.3023629143549159,
+      "grad_norm": 0.47212377938185984,
+      "learning_rate": 4.178724840208029e-06,
+      "loss": 0.1655,
+      "step": 1342
+    },
+    {
+      "epoch": 0.3025882220407244,
+      "grad_norm": 0.45808277304699035,
+      "learning_rate": 4.1773439407326474e-06,
+      "loss": 0.1606,
+      "step": 1343
+    },
+    {
+      "epoch": 0.3028135297265328,
+      "grad_norm": 0.4450063295423093,
+      "learning_rate": 4.175962109890697e-06,
+      "loss": 0.153,
+      "step": 1344
+    },
+    {
+      "epoch": 0.3030388374123412,
+      "grad_norm": 0.5067429640735599,
+      "learning_rate": 4.174579348449456e-06,
+      "loss": 0.1782,
+      "step": 1345
+    },
+    {
+      "epoch": 0.30326414509814964,
+      "grad_norm": 0.4744420430173419,
+      "learning_rate": 4.1731956571767215e-06,
+      "loss": 0.1586,
+      "step": 1346
+    },
+    {
+      "epoch": 0.3034894527839581,
+      "grad_norm": 0.4826024239192088,
+      "learning_rate": 4.171811036840805e-06,
+      "loss": 0.1614,
+      "step": 1347
+    },
+    {
+      "epoch": 0.30371476046976653,
+      "grad_norm": 0.47082954610521216,
+      "learning_rate": 4.170425488210534e-06,
+      "loss": 0.1706,
+      "step": 1348
+    },
+    {
+      "epoch": 0.30394006815557495,
+      "grad_norm": 0.4599812085595551,
+      "learning_rate": 4.169039012055255e-06,
+      "loss": 0.1622,
+      "step": 1349
+    },
+    {
+      "epoch": 0.30416537584138337,
+      "grad_norm": 0.4658111624747646,
+      "learning_rate": 4.167651609144822e-06,
+      "loss": 0.1649,
+      "step": 1350
+    },
+    {
+      "epoch": 0.30439068352719184,
+      "grad_norm": 0.48755223540020975,
+      "learning_rate": 4.166263280249613e-06,
+      "loss": 0.1669,
+      "step": 1351
+    },
+    {
+      "epoch": 0.30461599121300026,
+      "grad_norm": 0.45303825299908984,
+      "learning_rate": 4.164874026140511e-06,
+      "loss": 0.1643,
+      "step": 1352
+    },
+    {
+      "epoch": 0.3048412988988087,
+      "grad_norm": 0.445021146008286,
+      "learning_rate": 4.163483847588919e-06,
+      "loss": 0.1558,
+      "step": 1353
+    },
+    {
+      "epoch": 0.3050666065846171,
+      "grad_norm": 0.47655572636339527,
+      "learning_rate": 4.1620927453667515e-06,
+      "loss": 0.1687,
+      "step": 1354
+    },
+    {
+      "epoch": 0.30529191427042557,
+      "grad_norm": 0.47234459669526374,
+      "learning_rate": 4.160700720246435e-06,
+      "loss": 0.1628,
+      "step": 1355
+    },
+    {
+      "epoch": 0.305517221956234,
+      "grad_norm": 0.4531327995219235,
+      "learning_rate": 4.159307773000909e-06,
+      "loss": 0.1516,
+      "step": 1356
+    },
+    {
+      "epoch": 0.3057425296420424,
+      "grad_norm": 0.4992459885733053,
+      "learning_rate": 4.1579139044036265e-06,
+      "loss": 0.1661,
+      "step": 1357
+    },
+    {
+      "epoch": 0.3059678373278508,
+      "grad_norm": 0.45234523767743423,
+      "learning_rate": 4.15651911522855e-06,
+      "loss": 0.1591,
+      "step": 1358
+    },
+    {
+      "epoch": 0.3061931450136593,
+      "grad_norm": 0.5144568617456705,
+      "learning_rate": 4.155123406250153e-06,
+      "loss": 0.1842,
+      "step": 1359
+    },
+    {
+      "epoch": 0.3064184526994677,
+      "grad_norm": 0.4492467197923174,
+      "learning_rate": 4.153726778243422e-06,
+      "loss": 0.159,
+      "step": 1360
+    },
+    {
+      "epoch": 0.30664376038527613,
+      "grad_norm": 0.49125517977000377,
+      "learning_rate": 4.152329231983852e-06,
+      "loss": 0.1784,
+      "step": 1361
+    },
+    {
+      "epoch": 0.30686906807108455,
+      "grad_norm": 0.48383581152863314,
+      "learning_rate": 4.150930768247449e-06,
+      "loss": 0.1761,
+      "step": 1362
+    },
+    {
+      "epoch": 0.307094375756893,
+      "grad_norm": 0.45251816032706066,
+      "learning_rate": 4.149531387810727e-06,
+      "loss": 0.1717,
+      "step": 1363
+    },
+    {
+      "epoch": 0.30731968344270144,
+      "grad_norm": 0.4714618478306759,
+      "learning_rate": 4.148131091450709e-06,
+      "loss": 0.1601,
+      "step": 1364
+    },
+    {
+      "epoch": 0.30754499112850986,
+      "grad_norm": 0.46639711128092015,
+      "learning_rate": 4.14672987994493e-06,
+      "loss": 0.1492,
+      "step": 1365
+    },
+    {
+      "epoch": 0.3077702988143183,
+      "grad_norm": 0.5019168227780006,
+      "learning_rate": 4.145327754071427e-06,
+      "loss": 0.1785,
+      "step": 1366
+    },
+    {
+      "epoch": 0.30799560650012675,
+      "grad_norm": 0.5115191763440133,
+      "learning_rate": 4.1439247146087515e-06,
+      "loss": 0.1774,
+      "step": 1367
+    },
+    {
+      "epoch": 0.30822091418593517,
+      "grad_norm": 0.44762710081650325,
+      "learning_rate": 4.142520762335957e-06,
+      "loss": 0.1489,
+      "step": 1368
+    },
+    {
+      "epoch": 0.3084462218717436,
+      "grad_norm": 0.46505395959631984,
+      "learning_rate": 4.141115898032607e-06,
+      "loss": 0.1622,
+      "step": 1369
+    },
+    {
+      "epoch": 0.308671529557552,
+      "grad_norm": 0.49062895275209234,
+      "learning_rate": 4.13971012247877e-06,
+      "loss": 0.1589,
+      "step": 1370
+    },
+    {
+      "epoch": 0.3088968372433605,
+      "grad_norm": 0.4878995152656453,
+      "learning_rate": 4.138303436455019e-06,
+      "loss": 0.1713,
+      "step": 1371
+    },
+    {
+      "epoch": 0.3091221449291689,
+      "grad_norm": 0.4841704377642727,
+      "learning_rate": 4.136895840742437e-06,
+      "loss": 0.1707,
+      "step": 1372
+    },
+    {
+      "epoch": 0.3093474526149773,
+      "grad_norm": 0.46594485906792216,
+      "learning_rate": 4.1354873361226074e-06,
+      "loss": 0.1569,
+      "step": 1373
+    },
+    {
+      "epoch": 0.3095727603007858,
+      "grad_norm": 0.4799789377866047,
+      "learning_rate": 4.134077923377622e-06,
+      "loss": 0.1643,
+      "step": 1374
+    },
+    {
+      "epoch": 0.3097980679865942,
+      "grad_norm": 0.46024391817092386,
+      "learning_rate": 4.132667603290075e-06,
+      "loss": 0.175,
+      "step": 1375
+    },
+    {
+      "epoch": 0.3100233756724026,
+      "grad_norm": 0.5004998885019434,
+      "learning_rate": 4.131256376643062e-06,
+      "loss": 0.184,
+      "step": 1376
+    },
+    {
+      "epoch": 0.31024868335821104,
+      "grad_norm": 0.4496587322418268,
+      "learning_rate": 4.129844244220188e-06,
+      "loss": 0.1567,
+      "step": 1377
+    },
+    {
+      "epoch": 0.3104739910440195,
+      "grad_norm": 0.4381923644413625,
+      "learning_rate": 4.128431206805556e-06,
+      "loss": 0.1571,
+      "step": 1378
+    },
+    {
+      "epoch": 0.31069929872982793,
+      "grad_norm": 0.48281884874936964,
+      "learning_rate": 4.127017265183772e-06,
+      "loss": 0.1576,
+      "step": 1379
+    },
+    {
+      "epoch": 0.31092460641563635,
+      "grad_norm": 0.486608789161625,
+      "learning_rate": 4.125602420139947e-06,
+      "loss": 0.1736,
+      "step": 1380
+    },
+    {
+      "epoch": 0.31114991410144477,
+      "grad_norm": 0.43631321179674804,
+      "learning_rate": 4.124186672459691e-06,
+      "loss": 0.1521,
+      "step": 1381
+    },
+    {
+      "epoch": 0.31137522178725324,
+      "grad_norm": 0.46337301384693746,
+      "learning_rate": 4.122770022929114e-06,
+      "loss": 0.1693,
+      "step": 1382
+    },
+    {
+      "epoch": 0.31160052947306166,
+      "grad_norm": 0.49510374301430243,
+      "learning_rate": 4.121352472334832e-06,
+      "loss": 0.165,
+      "step": 1383
+    },
+    {
+      "epoch": 0.3118258371588701,
+      "grad_norm": 0.4898587365242441,
+      "learning_rate": 4.119934021463956e-06,
+      "loss": 0.1734,
+      "step": 1384
+    },
+    {
+      "epoch": 0.3120511448446785,
+      "grad_norm": 0.48861456754196575,
+      "learning_rate": 4.1185146711040995e-06,
+      "loss": 0.1704,
+      "step": 1385
+    },
+    {
+      "epoch": 0.31227645253048697,
+      "grad_norm": 0.49674442096608046,
+      "learning_rate": 4.117094422043374e-06,
+      "loss": 0.1752,
+      "step": 1386
+    },
+    {
+      "epoch": 0.3125017602162954,
+      "grad_norm": 0.5090139483439499,
+      "learning_rate": 4.115673275070392e-06,
+      "loss": 0.1706,
+      "step": 1387
+    },
+    {
+      "epoch": 0.3127270679021038,
+      "grad_norm": 0.46111045623593994,
+      "learning_rate": 4.114251230974263e-06,
+      "loss": 0.1633,
+      "step": 1388
+    },
+    {
+      "epoch": 0.3129523755879122,
+      "grad_norm": 0.4906796666813092,
+      "learning_rate": 4.1128282905445945e-06,
+      "loss": 0.1788,
+      "step": 1389
+    },
+    {
+      "epoch": 0.3131776832737207,
+      "grad_norm": 0.5216823967451348,
+      "learning_rate": 4.1114044545714935e-06,
+      "loss": 0.1779,
+      "step": 1390
+    },
+    {
+      "epoch": 0.3134029909595291,
+      "grad_norm": 0.4630187597122319,
+      "learning_rate": 4.1099797238455615e-06,
+      "loss": 0.175,
+      "step": 1391
+    },
+    {
+      "epoch": 0.31362829864533753,
+      "grad_norm": 0.4448150220936651,
+      "learning_rate": 4.1085540991579e-06,
+      "loss": 0.1599,
+      "step": 1392
+    },
+    {
+      "epoch": 0.31385360633114595,
+      "grad_norm": 0.47815315348196197,
+      "learning_rate": 4.107127581300105e-06,
+      "loss": 0.18,
+      "step": 1393
+    },
+    {
+      "epoch": 0.3140789140169544,
+      "grad_norm": 0.4647279014720571,
+      "learning_rate": 4.105700171064267e-06,
+      "loss": 0.1679,
+      "step": 1394
+    },
+    {
+      "epoch": 0.31430422170276284,
+      "grad_norm": 0.4755973884828333,
+      "learning_rate": 4.104271869242975e-06,
+      "loss": 0.1613,
+      "step": 1395
+    },
+    {
+      "epoch": 0.31452952938857126,
+      "grad_norm": 0.5158289979464339,
+      "learning_rate": 4.102842676629313e-06,
+      "loss": 0.1814,
+      "step": 1396
+    },
+    {
+      "epoch": 0.3147548370743797,
+      "grad_norm": 0.4612219449801974,
+      "learning_rate": 4.101412594016855e-06,
+      "loss": 0.1704,
+      "step": 1397
+    },
+    {
+      "epoch": 0.31498014476018815,
+      "grad_norm": 0.48199287409838215,
+      "learning_rate": 4.0999816221996755e-06,
+      "loss": 0.157,
+      "step": 1398
+    },
+    {
+      "epoch": 0.31520545244599657,
+      "grad_norm": 0.49588565495852915,
+      "learning_rate": 4.098549761972339e-06,
+      "loss": 0.1698,
+      "step": 1399
+    },
+    {
+      "epoch": 0.315430760131805,
+      "grad_norm": 0.47443029199155873,
+      "learning_rate": 4.097117014129903e-06,
+      "loss": 0.1705,
+      "step": 1400
+    },
+    {
+      "epoch": 0.3156560678176134,
+      "grad_norm": 0.4739011675953338,
+      "learning_rate": 4.095683379467922e-06,
+      "loss": 0.1746,
+      "step": 1401
+    },
+    {
+      "epoch": 0.3158813755034219,
+      "grad_norm": 0.5006985353128484,
+      "learning_rate": 4.094248858782436e-06,
+      "loss": 0.1749,
+      "step": 1402
+    },
+    {
+      "epoch": 0.3161066831892303,
+      "grad_norm": 0.48832615004361674,
+      "learning_rate": 4.092813452869983e-06,
+      "loss": 0.1748,
+      "step": 1403
+    },
+    {
+      "epoch": 0.3163319908750387,
+      "grad_norm": 0.5047270198243387,
+      "learning_rate": 4.091377162527592e-06,
+      "loss": 0.1741,
+      "step": 1404
+    },
+    {
+      "epoch": 0.31655729856084713,
+      "grad_norm": 0.48286040196834346,
+      "learning_rate": 4.089939988552778e-06,
+      "loss": 0.1703,
+      "step": 1405
+    },
+    {
+      "epoch": 0.3167826062466556,
+      "grad_norm": 0.4772214512873062,
+      "learning_rate": 4.088501931743551e-06,
+      "loss": 0.1844,
+      "step": 1406
+    },
+    {
+      "epoch": 0.317007913932464,
+      "grad_norm": 0.4700248977848288,
+      "learning_rate": 4.087062992898413e-06,
+      "loss": 0.1622,
+      "step": 1407
+    },
+    {
+      "epoch": 0.31723322161827244,
+      "grad_norm": 0.49138673902246965,
+      "learning_rate": 4.08562317281635e-06,
+      "loss": 0.1733,
+      "step": 1408
+    },
+    {
+      "epoch": 0.31745852930408086,
+      "grad_norm": 0.48634809473907503,
+      "learning_rate": 4.084182472296842e-06,
+      "loss": 0.171,
+      "step": 1409
+    },
+    {
+      "epoch": 0.31768383698988933,
+      "grad_norm": 0.4931411565996177,
+      "learning_rate": 4.082740892139856e-06,
+      "loss": 0.171,
+      "step": 1410
+    },
+    {
+      "epoch": 0.31790914467569775,
+      "grad_norm": 0.49274535939778674,
+      "learning_rate": 4.081298433145847e-06,
+      "loss": 0.1591,
+      "step": 1411
+    },
+    {
+      "epoch": 0.31813445236150617,
+      "grad_norm": 0.4698037285578985,
+      "learning_rate": 4.07985509611576e-06,
+      "loss": 0.1667,
+      "step": 1412
+    },
+    {
+      "epoch": 0.3183597600473146,
+      "grad_norm": 0.47696407295357496,
+      "learning_rate": 4.078410881851026e-06,
+      "loss": 0.1697,
+      "step": 1413
+    },
+    {
+      "epoch": 0.31858506773312306,
+      "grad_norm": 0.4827409751476058,
+      "learning_rate": 4.076965791153562e-06,
+      "loss": 0.1672,
+      "step": 1414
+    },
+    {
+      "epoch": 0.3188103754189315,
+      "grad_norm": 0.4841572963600362,
+      "learning_rate": 4.075519824825775e-06,
+      "loss": 0.1804,
+      "step": 1415
+    },
+    {
+      "epoch": 0.3190356831047399,
+      "grad_norm": 0.46260361476399064,
+      "learning_rate": 4.074072983670555e-06,
+      "loss": 0.17,
+      "step": 1416
+    },
+    {
+      "epoch": 0.3192609907905483,
+      "grad_norm": 0.4828365970593339,
+      "learning_rate": 4.072625268491279e-06,
+      "loss": 0.1716,
+      "step": 1417
+    },
+    {
+      "epoch": 0.3194862984763568,
+      "grad_norm": 0.5020748613802973,
+      "learning_rate": 4.071176680091809e-06,
+      "loss": 0.1864,
+      "step": 1418
+    },
+    {
+      "epoch": 0.3197116061621652,
+      "grad_norm": 0.46429356070269806,
+      "learning_rate": 4.069727219276493e-06,
+      "loss": 0.1533,
+      "step": 1419
+    },
+    {
+      "epoch": 0.3199369138479736,
+      "grad_norm": 0.47253365356597143,
+      "learning_rate": 4.068276886850162e-06,
+      "loss": 0.1539,
+      "step": 1420
+    },
+    {
+      "epoch": 0.3201622215337821,
+      "grad_norm": 0.43290691768683814,
+      "learning_rate": 4.066825683618132e-06,
+      "loss": 0.1523,
+      "step": 1421
+    },
+    {
+      "epoch": 0.3203875292195905,
+      "grad_norm": 0.5150603139209966,
+      "learning_rate": 4.065373610386201e-06,
+      "loss": 0.1906,
+      "step": 1422
+    },
+    {
+      "epoch": 0.32061283690539893,
+      "grad_norm": 0.4565348014998822,
+      "learning_rate": 4.063920667960652e-06,
+      "loss": 0.1645,
+      "step": 1423
+    },
+    {
+      "epoch": 0.32083814459120735,
+      "grad_norm": 0.4605096118081446,
+      "learning_rate": 4.06246685714825e-06,
+      "loss": 0.1629,
+      "step": 1424
+    },
+    {
+      "epoch": 0.3210634522770158,
+      "grad_norm": 0.4622701306481583,
+      "learning_rate": 4.061012178756242e-06,
+      "loss": 0.1683,
+      "step": 1425
+    },
+    {
+      "epoch": 0.32128875996282424,
+      "grad_norm": 0.4859787260584599,
+      "learning_rate": 4.059556633592356e-06,
+      "loss": 0.1594,
+      "step": 1426
+    },
+    {
+      "epoch": 0.32151406764863266,
+      "grad_norm": 0.48877370296912087,
+      "learning_rate": 4.058100222464802e-06,
+      "loss": 0.167,
+      "step": 1427
+    },
+    {
+      "epoch": 0.3217393753344411,
+      "grad_norm": 0.4627117579487038,
+      "learning_rate": 4.056642946182271e-06,
+      "loss": 0.1526,
+      "step": 1428
+    },
+    {
+      "epoch": 0.32196468302024955,
+      "grad_norm": 0.46613918156888673,
+      "learning_rate": 4.0551848055539345e-06,
+      "loss": 0.1653,
+      "step": 1429
+    },
+    {
+      "epoch": 0.32218999070605797,
+      "grad_norm": 0.503917741619968,
+      "learning_rate": 4.0537258013894434e-06,
+      "loss": 0.172,
+      "step": 1430
+    },
+    {
+      "epoch": 0.3224152983918664,
+      "grad_norm": 0.49156732022138344,
+      "learning_rate": 4.052265934498929e-06,
+      "loss": 0.1743,
+      "step": 1431
+    },
+    {
+      "epoch": 0.3226406060776748,
+      "grad_norm": 0.483390765528641,
+      "learning_rate": 4.0508052056929995e-06,
+      "loss": 0.158,
+      "step": 1432
+    },
+    {
+      "epoch": 0.3228659137634833,
+      "grad_norm": 0.46973338880450943,
+      "learning_rate": 4.049343615782744e-06,
+      "loss": 0.1647,
+      "step": 1433
+    },
+    {
+      "epoch": 0.3230912214492917,
+      "grad_norm": 0.47841122050317536,
+      "learning_rate": 4.047881165579729e-06,
+      "loss": 0.1714,
+      "step": 1434
+    },
+    {
+      "epoch": 0.3233165291351001,
+      "grad_norm": 0.4701686511398567,
+      "learning_rate": 4.046417855895999e-06,
+      "loss": 0.1657,
+      "step": 1435
+    },
+    {
+      "epoch": 0.32354183682090853,
+      "grad_norm": 0.4860143638189218,
+      "learning_rate": 4.044953687544074e-06,
+      "loss": 0.1715,
+      "step": 1436
+    },
+    {
+      "epoch": 0.323767144506717,
+      "grad_norm": 0.45709754277717596,
+      "learning_rate": 4.043488661336953e-06,
+      "loss": 0.1552,
+      "step": 1437
+    },
+    {
+      "epoch": 0.3239924521925254,
+      "grad_norm": 0.4679828926968299,
+      "learning_rate": 4.042022778088111e-06,
+      "loss": 0.1598,
+      "step": 1438
+    },
+    {
+      "epoch": 0.32421775987833384,
+      "grad_norm": 0.47861779407877914,
+      "learning_rate": 4.0405560386114975e-06,
+      "loss": 0.175,
+      "step": 1439
+    },
+    {
+      "epoch": 0.32444306756414226,
+      "grad_norm": 0.47207995414404774,
+      "learning_rate": 4.039088443721538e-06,
+      "loss": 0.1481,
+      "step": 1440
+    },
+    {
+      "epoch": 0.32466837524995074,
+      "grad_norm": 0.4701578331333167,
+      "learning_rate": 4.0376199942331335e-06,
+      "loss": 0.1621,
+      "step": 1441
+    },
+    {
+      "epoch": 0.32489368293575915,
+      "grad_norm": 0.46083892406658933,
+      "learning_rate": 4.03615069096166e-06,
+      "loss": 0.1662,
+      "step": 1442
+    },
+    {
+      "epoch": 0.32511899062156757,
+      "grad_norm": 0.4713751574792367,
+      "learning_rate": 4.034680534722966e-06,
+      "loss": 0.172,
+      "step": 1443
+    },
+    {
+      "epoch": 0.325344298307376,
+      "grad_norm": 0.4800419145752851,
+      "learning_rate": 4.033209526333375e-06,
+      "loss": 0.1635,
+      "step": 1444
+    },
+    {
+      "epoch": 0.32556960599318446,
+      "grad_norm": 0.5177479216025512,
+      "learning_rate": 4.0317376666096815e-06,
+      "loss": 0.1825,
+      "step": 1445
+    },
+    {
+      "epoch": 0.3257949136789929,
+      "grad_norm": 0.507169424745277,
+      "learning_rate": 4.030264956369158e-06,
+      "loss": 0.1634,
+      "step": 1446
+    },
+    {
+      "epoch": 0.3260202213648013,
+      "grad_norm": 0.49970993985240464,
+      "learning_rate": 4.028791396429541e-06,
+      "loss": 0.1706,
+      "step": 1447
+    },
+    {
+      "epoch": 0.3262455290506097,
+      "grad_norm": 0.48030204853136504,
+      "learning_rate": 4.0273169876090475e-06,
+      "loss": 0.1708,
+      "step": 1448
+    },
+    {
+      "epoch": 0.3264708367364182,
+      "grad_norm": 0.4610761076996909,
+      "learning_rate": 4.02584173072636e-06,
+      "loss": 0.158,
+      "step": 1449
+    },
+    {
+      "epoch": 0.3266961444222266,
+      "grad_norm": 0.464686959031091,
+      "learning_rate": 4.024365626600632e-06,
+      "loss": 0.1542,
+      "step": 1450
+    },
+    {
+      "epoch": 0.326921452108035,
+      "grad_norm": 0.48376124217372546,
+      "learning_rate": 4.022888676051492e-06,
+      "loss": 0.1625,
+      "step": 1451
+    },
+    {
+      "epoch": 0.32714675979384344,
+      "grad_norm": 0.4703117306092047,
+      "learning_rate": 4.021410879899035e-06,
+      "loss": 0.1669,
+      "step": 1452
+    },
+    {
+      "epoch": 0.3273720674796519,
+      "grad_norm": 0.48567988701291553,
+      "learning_rate": 4.019932238963824e-06,
+      "loss": 0.1627,
+      "step": 1453
+    },
+    {
+      "epoch": 0.32759737516546034,
+      "grad_norm": 0.47111752684595165,
+      "learning_rate": 4.018452754066895e-06,
+      "loss": 0.1675,
+      "step": 1454
+    },
+    {
+      "epoch": 0.32782268285126875,
+      "grad_norm": 0.46905414134840095,
+      "learning_rate": 4.016972426029751e-06,
+      "loss": 0.163,
+      "step": 1455
+    },
+    {
+      "epoch": 0.32804799053707717,
+      "grad_norm": 0.4647820038576828,
+      "learning_rate": 4.015491255674362e-06,
+      "loss": 0.1509,
+      "step": 1456
+    },
+    {
+      "epoch": 0.32827329822288565,
+      "grad_norm": 0.48517318183312064,
+      "learning_rate": 4.014009243823167e-06,
+      "loss": 0.178,
+      "step": 1457
+    },
+    {
+      "epoch": 0.32849860590869406,
+      "grad_norm": 0.4928940400289568,
+      "learning_rate": 4.012526391299073e-06,
+      "loss": 0.1638,
+      "step": 1458
+    },
+    {
+      "epoch": 0.3287239135945025,
+      "grad_norm": 0.48786731867399125,
+      "learning_rate": 4.01104269892545e-06,
+      "loss": 0.1738,
+      "step": 1459
+    },
+    {
+      "epoch": 0.3289492212803109,
+      "grad_norm": 0.49032486420130994,
+      "learning_rate": 4.0095581675261405e-06,
+      "loss": 0.1634,
+      "step": 1460
+    },
+    {
+      "epoch": 0.3291745289661194,
+      "grad_norm": 0.5219124278013928,
+      "learning_rate": 4.008072797925447e-06,
+      "loss": 0.172,
+      "step": 1461
+    },
+    {
+      "epoch": 0.3293998366519278,
+      "grad_norm": 0.45938422120363903,
+      "learning_rate": 4.006586590948141e-06,
+      "loss": 0.1567,
+      "step": 1462
+    },
+    {
+      "epoch": 0.3296251443377362,
+      "grad_norm": 0.49674977160069705,
+      "learning_rate": 4.005099547419458e-06,
+      "loss": 0.1807,
+      "step": 1463
+    },
+    {
+      "epoch": 0.3298504520235446,
+      "grad_norm": 0.49291053419468084,
+      "learning_rate": 4.003611668165097e-06,
+      "loss": 0.1635,
+      "step": 1464
+    },
+    {
+      "epoch": 0.3300757597093531,
+      "grad_norm": 0.47909951524830563,
+      "learning_rate": 4.0021229540112226e-06,
+      "loss": 0.1685,
+      "step": 1465
+    },
+    {
+      "epoch": 0.3303010673951615,
+      "grad_norm": 0.4967423779350966,
+      "learning_rate": 4.000633405784461e-06,
+      "loss": 0.1698,
+      "step": 1466
+    },
+    {
+      "epoch": 0.33052637508096994,
+      "grad_norm": 0.4947671017226202,
+      "learning_rate": 3.999143024311904e-06,
+      "loss": 0.1705,
+      "step": 1467
+    },
+    {
+      "epoch": 0.33075168276677835,
+      "grad_norm": 0.5001281128990965,
+      "learning_rate": 3.997651810421106e-06,
+      "loss": 0.1758,
+      "step": 1468
+    },
+    {
+      "epoch": 0.33097699045258683,
+      "grad_norm": 0.47166189147867515,
+      "learning_rate": 3.99615976494008e-06,
+      "loss": 0.1621,
+      "step": 1469
+    },
+    {
+      "epoch": 0.33120229813839525,
+      "grad_norm": 0.5142434013369097,
+      "learning_rate": 3.994666888697304e-06,
+      "loss": 0.1845,
+      "step": 1470
+    },
+    {
+      "epoch": 0.33142760582420366,
+      "grad_norm": 0.4805539617231105,
+      "learning_rate": 3.993173182521718e-06,
+      "loss": 0.1605,
+      "step": 1471
+    },
+    {
+      "epoch": 0.33165291351001214,
+      "grad_norm": 0.5043483938154787,
+      "learning_rate": 3.991678647242719e-06,
+      "loss": 0.175,
+      "step": 1472
+    },
+    {
+      "epoch": 0.33187822119582056,
+      "grad_norm": 0.478038959369919,
+      "learning_rate": 3.990183283690169e-06,
+      "loss": 0.1596,
+      "step": 1473
+    },
+    {
+      "epoch": 0.332103528881629,
+      "grad_norm": 0.4841686206196171,
+      "learning_rate": 3.988687092694386e-06,
+      "loss": 0.1647,
+      "step": 1474
+    },
+    {
+      "epoch": 0.3323288365674374,
+      "grad_norm": 0.4846814128855193,
+      "learning_rate": 3.98719007508615e-06,
+      "loss": 0.1702,
+      "step": 1475
+    },
+    {
+      "epoch": 0.33255414425324586,
+      "grad_norm": 0.4858363667117268,
+      "learning_rate": 3.985692231696699e-06,
+      "loss": 0.1689,
+      "step": 1476
+    },
+    {
+      "epoch": 0.3327794519390543,
+      "grad_norm": 0.48567860031406734,
+      "learning_rate": 3.98419356335773e-06,
+      "loss": 0.1694,
+      "step": 1477
+    },
+    {
+      "epoch": 0.3330047596248627,
+      "grad_norm": 0.49126963786538463,
+      "learning_rate": 3.982694070901396e-06,
+      "loss": 0.1714,
+      "step": 1478
+    },
+    {
+      "epoch": 0.3332300673106711,
+      "grad_norm": 0.46435208474415085,
+      "learning_rate": 3.981193755160311e-06,
+      "loss": 0.1632,
+      "step": 1479
+    },
+    {
+      "epoch": 0.3334553749964796,
+      "grad_norm": 0.5103818515068653,
+      "learning_rate": 3.979692616967543e-06,
+      "loss": 0.1906,
+      "step": 1480
+    },
+    {
+      "epoch": 0.333680682682288,
+      "grad_norm": 0.48145456741346393,
+      "learning_rate": 3.9781906571566195e-06,
+      "loss": 0.1649,
+      "step": 1481
+    },
+    {
+      "epoch": 0.33390599036809643,
+      "grad_norm": 0.4853961796628509,
+      "learning_rate": 3.976687876561523e-06,
+      "loss": 0.1774,
+      "step": 1482
+    },
+    {
+      "epoch": 0.33413129805390485,
+      "grad_norm": 0.4799122387066231,
+      "learning_rate": 3.975184276016689e-06,
+      "loss": 0.1739,
+      "step": 1483
+    },
+    {
+      "epoch": 0.3343566057397133,
+      "grad_norm": 0.46563910497802685,
+      "learning_rate": 3.973679856357014e-06,
+      "loss": 0.1554,
+      "step": 1484
+    },
+    {
+      "epoch": 0.33458191342552174,
+      "grad_norm": 0.44840230729722863,
+      "learning_rate": 3.972174618417843e-06,
+      "loss": 0.1588,
+      "step": 1485
+    },
+    {
+      "epoch": 0.33480722111133016,
+      "grad_norm": 0.505445345978517,
+      "learning_rate": 3.970668563034982e-06,
+      "loss": 0.1753,
+      "step": 1486
+    },
+    {
+      "epoch": 0.3350325287971386,
+      "grad_norm": 0.4995685182105176,
+      "learning_rate": 3.9691616910446845e-06,
+      "loss": 0.1701,
+      "step": 1487
+    },
+    {
+      "epoch": 0.33525783648294705,
+      "grad_norm": 0.509006048277001,
+      "learning_rate": 3.967654003283662e-06,
+      "loss": 0.1723,
+      "step": 1488
+    },
+    {
+      "epoch": 0.33548314416875546,
+      "grad_norm": 0.5169005973891307,
+      "learning_rate": 3.966145500589076e-06,
+      "loss": 0.1803,
+      "step": 1489
+    },
+    {
+      "epoch": 0.3357084518545639,
+      "grad_norm": 0.5079388196082648,
+      "learning_rate": 3.9646361837985435e-06,
+      "loss": 0.1791,
+      "step": 1490
+    },
+    {
+      "epoch": 0.3359337595403723,
+      "grad_norm": 0.4560258431578383,
+      "learning_rate": 3.9631260537501304e-06,
+      "loss": 0.1524,
+      "step": 1491
+    },
+    {
+      "epoch": 0.3361590672261808,
+      "grad_norm": 0.4766011031285634,
+      "learning_rate": 3.961615111282357e-06,
+      "loss": 0.1781,
+      "step": 1492
+    },
+    {
+      "epoch": 0.3363843749119892,
+      "grad_norm": 0.4567234424418434,
+      "learning_rate": 3.960103357234192e-06,
+      "loss": 0.1643,
+      "step": 1493
+    },
+    {
+      "epoch": 0.3366096825977976,
+      "grad_norm": 0.49289766811652796,
+      "learning_rate": 3.958590792445057e-06,
+      "loss": 0.1823,
+      "step": 1494
+    },
+    {
+      "epoch": 0.33683499028360603,
+      "grad_norm": 0.4848856116585081,
+      "learning_rate": 3.957077417754822e-06,
+      "loss": 0.166,
+      "step": 1495
+    },
+    {
+      "epoch": 0.3370602979694145,
+      "grad_norm": 0.4679747882685483,
+      "learning_rate": 3.9555632340038075e-06,
+      "loss": 0.1774,
+      "step": 1496
+    },
+    {
+      "epoch": 0.3372856056552229,
+      "grad_norm": 0.48350839858166583,
+      "learning_rate": 3.9540482420327845e-06,
+      "loss": 0.1725,
+      "step": 1497
+    },
+    {
+      "epoch": 0.33751091334103134,
+      "grad_norm": 0.4863933008059015,
+      "learning_rate": 3.9525324426829716e-06,
+      "loss": 0.1641,
+      "step": 1498
+    },
+    {
+      "epoch": 0.33773622102683976,
+      "grad_norm": 0.46827754368039665,
+      "learning_rate": 3.951015836796034e-06,
+      "loss": 0.1524,
+      "step": 1499
+    },
+    {
+      "epoch": 0.33796152871264823,
+      "grad_norm": 0.44169102557800627,
+      "learning_rate": 3.949498425214088e-06,
+      "loss": 0.1513,
+      "step": 1500
+    },
+    {
+      "epoch": 0.33796152871264823,
+      "eval_loss": 0.16632719337940216,
+      "eval_runtime": 56.8474,
+      "eval_samples_per_second": 50.486,
+      "eval_steps_per_second": 6.315,
+      "step": 1500
+    },
+    {
+      "epoch": 0.33818683639845665,
+      "grad_norm": 0.4878295448211993,
+      "learning_rate": 3.947980208779693e-06,
+      "loss": 0.1653,
+      "step": 1501
+    },
+    {
+      "epoch": 0.33841214408426507,
+      "grad_norm": 0.49941504395630504,
+      "learning_rate": 3.946461188335863e-06,
+      "loss": 0.1827,
+      "step": 1502
+    },
+    {
+      "epoch": 0.3386374517700735,
+      "grad_norm": 0.4870788561570751,
+      "learning_rate": 3.944941364726049e-06,
+      "loss": 0.1946,
+      "step": 1503
+    },
+    {
+      "epoch": 0.33886275945588196,
+      "grad_norm": 0.46386674326836835,
+      "learning_rate": 3.943420738794153e-06,
+      "loss": 0.1597,
+      "step": 1504
+    },
+    {
+      "epoch": 0.3390880671416904,
+      "grad_norm": 0.5153906687266364,
+      "learning_rate": 3.941899311384525e-06,
+      "loss": 0.1768,
+      "step": 1505
+    },
+    {
+      "epoch": 0.3393133748274988,
+      "grad_norm": 0.527331696836763,
+      "learning_rate": 3.9403770833419535e-06,
+      "loss": 0.1707,
+      "step": 1506
+    },
+    {
+      "epoch": 0.3395386825133072,
+      "grad_norm": 0.4549856824962382,
+      "learning_rate": 3.938854055511676e-06,
+      "loss": 0.1529,
+      "step": 1507
+    },
+    {
+      "epoch": 0.3397639901991157,
+      "grad_norm": 0.4933487941824393,
+      "learning_rate": 3.937330228739374e-06,
+      "loss": 0.1702,
+      "step": 1508
+    },
+    {
+      "epoch": 0.3399892978849241,
+      "grad_norm": 0.4879928825898799,
+      "learning_rate": 3.9358056038711714e-06,
+      "loss": 0.1722,
+      "step": 1509
+    },
+    {
+      "epoch": 0.3402146055707325,
+      "grad_norm": 0.4636410414321768,
+      "learning_rate": 3.934280181753634e-06,
+      "loss": 0.1651,
+      "step": 1510
+    },
+    {
+      "epoch": 0.34043991325654094,
+      "grad_norm": 0.45711139575548393,
+      "learning_rate": 3.932753963233773e-06,
+      "loss": 0.1463,
+      "step": 1511
+    },
+    {
+      "epoch": 0.3406652209423494,
+      "grad_norm": 0.46149943138988064,
+      "learning_rate": 3.931226949159041e-06,
+      "loss": 0.1645,
+      "step": 1512
+    },
+    {
+      "epoch": 0.34089052862815783,
+      "grad_norm": 0.48167999322238986,
+      "learning_rate": 3.9296991403773325e-06,
+      "loss": 0.1666,
+      "step": 1513
+    },
+    {
+      "epoch": 0.34111583631396625,
+      "grad_norm": 0.4665831445664159,
+      "learning_rate": 3.9281705377369814e-06,
+      "loss": 0.149,
+      "step": 1514
+    },
+    {
+      "epoch": 0.34134114399977467,
+      "grad_norm": 0.45575953209532205,
+      "learning_rate": 3.9266411420867635e-06,
+      "loss": 0.147,
+      "step": 1515
+    },
+    {
+      "epoch": 0.34156645168558314,
+      "grad_norm": 0.5058205173300985,
+      "learning_rate": 3.925110954275897e-06,
+      "loss": 0.1681,
+      "step": 1516
+    },
+    {
+      "epoch": 0.34179175937139156,
+      "grad_norm": 0.4522501093222901,
+      "learning_rate": 3.923579975154037e-06,
+      "loss": 0.1587,
+      "step": 1517
+    },
+    {
+      "epoch": 0.3420170670572,
+      "grad_norm": 0.479996247672822,
+      "learning_rate": 3.922048205571279e-06,
+      "loss": 0.1718,
+      "step": 1518
+    },
+    {
+      "epoch": 0.34224237474300845,
+      "grad_norm": 0.4954333978709633,
+      "learning_rate": 3.920515646378159e-06,
+      "loss": 0.1648,
+      "step": 1519
+    },
+    {
+      "epoch": 0.34246768242881687,
+      "grad_norm": 0.4822288274252376,
+      "learning_rate": 3.918982298425647e-06,
+      "loss": 0.1799,
+      "step": 1520
+    },
+    {
+      "epoch": 0.3426929901146253,
+      "grad_norm": 0.46648553853872377,
+      "learning_rate": 3.917448162565157e-06,
+      "loss": 0.1433,
+      "step": 1521
+    },
+    {
+      "epoch": 0.3429182978004337,
+      "grad_norm": 0.47922896446110114,
+      "learning_rate": 3.915913239648535e-06,
+      "loss": 0.1742,
+      "step": 1522
+    },
+    {
+      "epoch": 0.3431436054862422,
+      "grad_norm": 0.4812863396173565,
+      "learning_rate": 3.91437753052807e-06,
+      "loss": 0.1707,
+      "step": 1523
+    },
+    {
+      "epoch": 0.3433689131720506,
+      "grad_norm": 0.45912938511177226,
+      "learning_rate": 3.91284103605648e-06,
+      "loss": 0.1573,
+      "step": 1524
+    },
+    {
+      "epoch": 0.343594220857859,
+      "grad_norm": 0.4950628079069977,
+      "learning_rate": 3.911303757086925e-06,
+      "loss": 0.1679,
+      "step": 1525
+    },
+    {
+      "epoch": 0.34381952854366743,
+      "grad_norm": 0.4467689502898321,
+      "learning_rate": 3.909765694473e-06,
+      "loss": 0.1597,
+      "step": 1526
+    },
+    {
+      "epoch": 0.3440448362294759,
+      "grad_norm": 0.5129219990156939,
+      "learning_rate": 3.908226849068731e-06,
+      "loss": 0.1661,
+      "step": 1527
+    },
+    {
+      "epoch": 0.3442701439152843,
+      "grad_norm": 0.4551952170221029,
+      "learning_rate": 3.906687221728583e-06,
+      "loss": 0.1643,
+      "step": 1528
+    },
+    {
+      "epoch": 0.34449545160109274,
+      "grad_norm": 0.47170480347901833,
+      "learning_rate": 3.905146813307455e-06,
+      "loss": 0.1611,
+      "step": 1529
+    },
+    {
+      "epoch": 0.34472075928690116,
+      "grad_norm": 0.48001748868857014,
+      "learning_rate": 3.903605624660676e-06,
+      "loss": 0.1612,
+      "step": 1530
+    },
+    {
+      "epoch": 0.34494606697270963,
+      "grad_norm": 0.4748628275155922,
+      "learning_rate": 3.902063656644012e-06,
+      "loss": 0.1631,
+      "step": 1531
+    },
+    {
+      "epoch": 0.34517137465851805,
+      "grad_norm": 0.5139859239848883,
+      "learning_rate": 3.900520910113659e-06,
+      "loss": 0.1914,
+      "step": 1532
+    },
+    {
+      "epoch": 0.34539668234432647,
+      "grad_norm": 0.4878759796792643,
+      "learning_rate": 3.898977385926249e-06,
+      "loss": 0.1718,
+      "step": 1533
+    },
+    {
+      "epoch": 0.3456219900301349,
+      "grad_norm": 0.4716708640496516,
+      "learning_rate": 3.897433084938841e-06,
+      "loss": 0.1526,
+      "step": 1534
+    },
+    {
+      "epoch": 0.34584729771594336,
+      "grad_norm": 0.46184841826766815,
+      "learning_rate": 3.895888008008929e-06,
+      "loss": 0.1654,
+      "step": 1535
+    },
+    {
+      "epoch": 0.3460726054017518,
+      "grad_norm": 0.5164135232993046,
+      "learning_rate": 3.894342155994437e-06,
+      "loss": 0.1811,
+      "step": 1536
+    },
+    {
+      "epoch": 0.3462979130875602,
+      "grad_norm": 0.48767451819098034,
+      "learning_rate": 3.892795529753718e-06,
+      "loss": 0.1667,
+      "step": 1537
+    },
+    {
+      "epoch": 0.3465232207733686,
+      "grad_norm": 0.4712763769177538,
+      "learning_rate": 3.891248130145556e-06,
+      "loss": 0.1625,
+      "step": 1538
+    },
+    {
+      "epoch": 0.3467485284591771,
+      "grad_norm": 0.4849166281449554,
+      "learning_rate": 3.889699958029166e-06,
+      "loss": 0.1593,
+      "step": 1539
+    },
+    {
+      "epoch": 0.3469738361449855,
+      "grad_norm": 0.4780505339438159,
+      "learning_rate": 3.888151014264189e-06,
+      "loss": 0.1652,
+      "step": 1540
+    },
+    {
+      "epoch": 0.3471991438307939,
+      "grad_norm": 0.4459016280706361,
+      "learning_rate": 3.886601299710694e-06,
+      "loss": 0.1517,
+      "step": 1541
+    },
+    {
+      "epoch": 0.34742445151660234,
+      "grad_norm": 0.45591641194578175,
+      "learning_rate": 3.885050815229182e-06,
+      "loss": 0.1511,
+      "step": 1542
+    },
+    {
+      "epoch": 0.3476497592024108,
+      "grad_norm": 0.4832877784381766,
+      "learning_rate": 3.88349956168058e-06,
+      "loss": 0.1684,
+      "step": 1543
+    },
+    {
+      "epoch": 0.34787506688821923,
+      "grad_norm": 0.4986224178690113,
+      "learning_rate": 3.881947539926239e-06,
+      "loss": 0.1691,
+      "step": 1544
+    },
+    {
+      "epoch": 0.34810037457402765,
+      "grad_norm": 0.4775733324064565,
+      "learning_rate": 3.880394750827939e-06,
+      "loss": 0.1615,
+      "step": 1545
+    },
+    {
+      "epoch": 0.34832568225983607,
+      "grad_norm": 0.4858017668488311,
+      "learning_rate": 3.878841195247888e-06,
+      "loss": 0.1811,
+      "step": 1546
+    },
+    {
+      "epoch": 0.34855098994564454,
+      "grad_norm": 0.48153045793870897,
+      "learning_rate": 3.877286874048716e-06,
+      "loss": 0.1718,
+      "step": 1547
+    },
+    {
+      "epoch": 0.34877629763145296,
+      "grad_norm": 0.512083637621429,
+      "learning_rate": 3.875731788093478e-06,
+      "loss": 0.1759,
+      "step": 1548
+    },
+    {
+      "epoch": 0.3490016053172614,
+      "grad_norm": 0.47558354766986316,
+      "learning_rate": 3.874175938245659e-06,
+      "loss": 0.1611,
+      "step": 1549
+    },
+    {
+      "epoch": 0.3492269130030698,
+      "grad_norm": 0.47594444437190775,
+      "learning_rate": 3.872619325369162e-06,
+      "loss": 0.1642,
+      "step": 1550
+    },
+    {
+      "epoch": 0.34945222068887827,
+      "grad_norm": 0.49698776261142624,
+      "learning_rate": 3.871061950328317e-06,
+      "loss": 0.1647,
+      "step": 1551
+    },
+    {
+      "epoch": 0.3496775283746867,
+      "grad_norm": 0.5120833541363724,
+      "learning_rate": 3.869503813987876e-06,
+      "loss": 0.1654,
+      "step": 1552
+    },
+    {
+      "epoch": 0.3499028360604951,
+      "grad_norm": 0.5070864621540254,
+      "learning_rate": 3.867944917213014e-06,
+      "loss": 0.1703,
+      "step": 1553
+    },
+    {
+      "epoch": 0.3501281437463035,
+      "grad_norm": 0.45295202225033465,
+      "learning_rate": 3.866385260869327e-06,
+      "loss": 0.1448,
+      "step": 1554
+    },
+    {
+      "epoch": 0.350353451432112,
+      "grad_norm": 0.4918072139538417,
+      "learning_rate": 3.864824845822837e-06,
+      "loss": 0.1721,
+      "step": 1555
+    },
+    {
+      "epoch": 0.3505787591179204,
+      "grad_norm": 0.5257156167287909,
+      "learning_rate": 3.8632636729399815e-06,
+      "loss": 0.1758,
+      "step": 1556
+    },
+    {
+      "epoch": 0.35080406680372883,
+      "grad_norm": 0.4680727482304966,
+      "learning_rate": 3.861701743087622e-06,
+      "loss": 0.164,
+      "step": 1557
+    },
+    {
+      "epoch": 0.35102937448953725,
+      "grad_norm": 0.4875956186422521,
+      "learning_rate": 3.860139057133042e-06,
+      "loss": 0.1768,
+      "step": 1558
+    },
+    {
+      "epoch": 0.3512546821753457,
+      "grad_norm": 0.47233606422385976,
+      "learning_rate": 3.858575615943941e-06,
+      "loss": 0.1698,
+      "step": 1559
+    },
+    {
+      "epoch": 0.35147998986115414,
+      "grad_norm": 0.5025916927692134,
+      "learning_rate": 3.85701142038844e-06,
+      "loss": 0.1854,
+      "step": 1560
+    },
+    {
+      "epoch": 0.35170529754696256,
+      "grad_norm": 0.4746547374579445,
+      "learning_rate": 3.855446471335078e-06,
+      "loss": 0.1572,
+      "step": 1561
+    },
+    {
+      "epoch": 0.351930605232771,
+      "grad_norm": 0.4794744494656748,
+      "learning_rate": 3.853880769652815e-06,
+      "loss": 0.1691,
+      "step": 1562
+    },
+    {
+      "epoch": 0.35215591291857945,
+      "grad_norm": 0.4843099574697001,
+      "learning_rate": 3.852314316211023e-06,
+      "loss": 0.1672,
+      "step": 1563
+    },
+    {
+      "epoch": 0.35238122060438787,
+      "grad_norm": 0.48546921942948973,
+      "learning_rate": 3.850747111879499e-06,
+      "loss": 0.1786,
+      "step": 1564
+    },
+    {
+      "epoch": 0.3526065282901963,
+      "grad_norm": 0.45578063099662863,
+      "learning_rate": 3.84917915752845e-06,
+      "loss": 0.1571,
+      "step": 1565
+    },
+    {
+      "epoch": 0.3528318359760047,
+      "grad_norm": 0.469326755886643,
+      "learning_rate": 3.8476104540285054e-06,
+      "loss": 0.1652,
+      "step": 1566
+    },
+    {
+      "epoch": 0.3530571436618132,
+      "grad_norm": 0.533642189695708,
+      "learning_rate": 3.846041002250705e-06,
+      "loss": 0.1787,
+      "step": 1567
+    },
+    {
+      "epoch": 0.3532824513476216,
+      "grad_norm": 0.4602243037230457,
+      "learning_rate": 3.84447080306651e-06,
+      "loss": 0.1672,
+      "step": 1568
+    },
+    {
+      "epoch": 0.35350775903343,
+      "grad_norm": 0.4914227338324527,
+      "learning_rate": 3.842899857347792e-06,
+      "loss": 0.1731,
+      "step": 1569
+    },
+    {
+      "epoch": 0.3537330667192385,
+      "grad_norm": 0.4528175947193335,
+      "learning_rate": 3.841328165966837e-06,
+      "loss": 0.1582,
+      "step": 1570
+    },
+    {
+      "epoch": 0.3539583744050469,
+      "grad_norm": 0.5043477043072159,
+      "learning_rate": 3.839755729796349e-06,
+      "loss": 0.1756,
+      "step": 1571
+    },
+    {
+      "epoch": 0.3541836820908553,
+      "grad_norm": 0.45252198647800007,
+      "learning_rate": 3.838182549709442e-06,
+      "loss": 0.1542,
+      "step": 1572
+    },
+    {
+      "epoch": 0.35440898977666374,
+      "grad_norm": 0.5153655969655626,
+      "learning_rate": 3.8366086265796445e-06,
+      "loss": 0.1788,
+      "step": 1573
+    },
+    {
+      "epoch": 0.3546342974624722,
+      "grad_norm": 0.46848812709265947,
+      "learning_rate": 3.835033961280898e-06,
+      "loss": 0.1608,
+      "step": 1574
+    },
+    {
+      "epoch": 0.35485960514828063,
+      "grad_norm": 0.49700435600318305,
+      "learning_rate": 3.8334585546875544e-06,
+      "loss": 0.1628,
+      "step": 1575
+    },
+    {
+      "epoch": 0.35508491283408905,
+      "grad_norm": 0.4515771767768039,
+      "learning_rate": 3.831882407674379e-06,
+      "loss": 0.1454,
+      "step": 1576
+    },
+    {
+      "epoch": 0.35531022051989747,
+      "grad_norm": 0.46810553468047617,
+      "learning_rate": 3.830305521116546e-06,
+      "loss": 0.1479,
+      "step": 1577
+    },
+    {
+      "epoch": 0.35553552820570594,
+      "grad_norm": 0.5185799810546041,
+      "learning_rate": 3.828727895889644e-06,
+      "loss": 0.1867,
+      "step": 1578
+    },
+    {
+      "epoch": 0.35576083589151436,
+      "grad_norm": 0.4648477693043239,
+      "learning_rate": 3.827149532869668e-06,
+      "loss": 0.1632,
+      "step": 1579
+    },
+    {
+      "epoch": 0.3559861435773228,
+      "grad_norm": 0.548001166651646,
+      "learning_rate": 3.825570432933026e-06,
+      "loss": 0.1737,
+      "step": 1580
+    },
+    {
+      "epoch": 0.3562114512631312,
+      "grad_norm": 0.5151582597577937,
+      "learning_rate": 3.823990596956531e-06,
+      "loss": 0.1665,
+      "step": 1581
+    },
+    {
+      "epoch": 0.35643675894893967,
+      "grad_norm": 0.46553401719297743,
+      "learning_rate": 3.8224100258174066e-06,
+      "loss": 0.1617,
+      "step": 1582
+    },
+    {
+      "epoch": 0.3566620666347481,
+      "grad_norm": 0.48246926237586046,
+      "learning_rate": 3.820828720393287e-06,
+      "loss": 0.1705,
+      "step": 1583
+    },
+    {
+      "epoch": 0.3568873743205565,
+      "grad_norm": 0.4609227338672053,
+      "learning_rate": 3.819246681562212e-06,
+      "loss": 0.1542,
+      "step": 1584
+    },
+    {
+      "epoch": 0.3571126820063649,
+      "grad_norm": 0.4806312081628679,
+      "learning_rate": 3.817663910202628e-06,
+      "loss": 0.1545,
+      "step": 1585
+    },
+    {
+      "epoch": 0.3573379896921734,
+      "grad_norm": 0.4522822568915773,
+      "learning_rate": 3.81608040719339e-06,
+      "loss": 0.1584,
+      "step": 1586
+    },
+    {
+      "epoch": 0.3575632973779818,
+      "grad_norm": 0.4831632255864022,
+      "learning_rate": 3.8144961734137566e-06,
+      "loss": 0.1617,
+      "step": 1587
+    },
+    {
+      "epoch": 0.35778860506379023,
+      "grad_norm": 0.46465414372602604,
+      "learning_rate": 3.812911209743395e-06,
+      "loss": 0.1448,
+      "step": 1588
+    },
+    {
+      "epoch": 0.35801391274959865,
+      "grad_norm": 0.47415748028404553,
+      "learning_rate": 3.8113255170623763e-06,
+      "loss": 0.1504,
+      "step": 1589
+    },
+    {
+      "epoch": 0.3582392204354071,
+      "grad_norm": 0.49482383184695794,
+      "learning_rate": 3.809739096251176e-06,
+      "loss": 0.1716,
+      "step": 1590
+    },
+    {
+      "epoch": 0.35846452812121554,
+      "grad_norm": 0.4760533886621652,
+      "learning_rate": 3.8081519481906747e-06,
+      "loss": 0.1712,
+      "step": 1591
+    },
+    {
+      "epoch": 0.35868983580702396,
+      "grad_norm": 0.4693732270317497,
+      "learning_rate": 3.8065640737621566e-06,
+      "loss": 0.1579,
+      "step": 1592
+    },
+    {
+      "epoch": 0.3589151434928324,
+      "grad_norm": 0.49273079763166744,
+      "learning_rate": 3.804975473847309e-06,
+      "loss": 0.1643,
+      "step": 1593
+    },
+    {
+      "epoch": 0.35914045117864085,
+      "grad_norm": 0.48411925790937316,
+      "learning_rate": 3.803386149328223e-06,
+      "loss": 0.1762,
+      "step": 1594
+    },
+    {
+      "epoch": 0.35936575886444927,
+      "grad_norm": 0.5103521369943442,
+      "learning_rate": 3.8017961010873904e-06,
+      "loss": 0.1633,
+      "step": 1595
+    },
+    {
+      "epoch": 0.3595910665502577,
+      "grad_norm": 0.4853884510594153,
+      "learning_rate": 3.8002053300077056e-06,
+      "loss": 0.1827,
+      "step": 1596
+    },
+    {
+      "epoch": 0.3598163742360661,
+      "grad_norm": 0.48135448571531514,
+      "learning_rate": 3.7986138369724664e-06,
+      "loss": 0.166,
+      "step": 1597
+    },
+    {
+      "epoch": 0.3600416819218746,
+      "grad_norm": 0.5077422631864816,
+      "learning_rate": 3.7970216228653667e-06,
+      "loss": 0.164,
+      "step": 1598
+    },
+    {
+      "epoch": 0.360266989607683,
+      "grad_norm": 0.4604975406803662,
+      "learning_rate": 3.795428688570505e-06,
+      "loss": 0.1577,
+      "step": 1599
+    },
+    {
+      "epoch": 0.3604922972934914,
+      "grad_norm": 0.5169319139843352,
+      "learning_rate": 3.7938350349723784e-06,
+      "loss": 0.1734,
+      "step": 1600
+    },
+    {
+      "epoch": 0.36071760497929983,
+      "grad_norm": 0.48315257961110103,
+      "learning_rate": 3.792240662955884e-06,
+      "loss": 0.1683,
+      "step": 1601
+    },
+    {
+      "epoch": 0.3609429126651083,
+      "grad_norm": 0.4443054345156236,
+      "learning_rate": 3.7906455734063156e-06,
+      "loss": 0.1583,
+      "step": 1602
+    },
+    {
+      "epoch": 0.3611682203509167,
+      "grad_norm": 0.5015309298690244,
+      "learning_rate": 3.7890497672093686e-06,
+      "loss": 0.1759,
+      "step": 1603
+    },
+    {
+      "epoch": 0.36139352803672514,
+      "grad_norm": 0.4893223920945183,
+      "learning_rate": 3.7874532452511324e-06,
+      "loss": 0.1594,
+      "step": 1604
+    },
+    {
+      "epoch": 0.36161883572253356,
+      "grad_norm": 0.4687661324719426,
+      "learning_rate": 3.785856008418099e-06,
+      "loss": 0.1646,
+      "step": 1605
+    },
+    {
+      "epoch": 0.36184414340834203,
+      "grad_norm": 0.4862531688588126,
+      "learning_rate": 3.7842580575971533e-06,
+      "loss": 0.1724,
+      "step": 1606
+    },
+    {
+      "epoch": 0.36206945109415045,
+      "grad_norm": 0.508026435849886,
+      "learning_rate": 3.782659393675577e-06,
+      "loss": 0.1673,
+      "step": 1607
+    },
+    {
+      "epoch": 0.36229475877995887,
+      "grad_norm": 0.4917181311635296,
+      "learning_rate": 3.7810600175410493e-06,
+      "loss": 0.1503,
+      "step": 1608
+    },
+    {
+      "epoch": 0.3625200664657673,
+      "grad_norm": 0.47047467772282814,
+      "learning_rate": 3.7794599300816435e-06,
+      "loss": 0.1601,
+      "step": 1609
+    },
+    {
+      "epoch": 0.36274537415157576,
+      "grad_norm": 0.48120667803642914,
+      "learning_rate": 3.77785913218583e-06,
+      "loss": 0.1767,
+      "step": 1610
+    },
+    {
+      "epoch": 0.3629706818373842,
+      "grad_norm": 0.47399195684134654,
+      "learning_rate": 3.7762576247424707e-06,
+      "loss": 0.1625,
+      "step": 1611
+    },
+    {
+      "epoch": 0.3631959895231926,
+      "grad_norm": 0.4634113964733802,
+      "learning_rate": 3.7746554086408245e-06,
+      "loss": 0.1531,
+      "step": 1612
+    },
+    {
+      "epoch": 0.363421297209001,
+      "grad_norm": 0.442021039632381,
+      "learning_rate": 3.7730524847705407e-06,
+      "loss": 0.1638,
+      "step": 1613
+    },
+    {
+      "epoch": 0.3636466048948095,
+      "grad_norm": 0.48085227932801655,
+      "learning_rate": 3.7714488540216637e-06,
+      "loss": 0.1649,
+      "step": 1614
+    },
+    {
+      "epoch": 0.3638719125806179,
+      "grad_norm": 0.4517008974302194,
+      "learning_rate": 3.7698445172846305e-06,
+      "loss": 0.153,
+      "step": 1615
+    },
+    {
+      "epoch": 0.3640972202664263,
+      "grad_norm": 0.4824941635671589,
+      "learning_rate": 3.7682394754502687e-06,
+      "loss": 0.168,
+      "step": 1616
+    },
+    {
+      "epoch": 0.3643225279522348,
+      "grad_norm": 0.481444912322331,
+      "learning_rate": 3.7666337294097987e-06,
+      "loss": 0.1614,
+      "step": 1617
+    },
+    {
+      "epoch": 0.3645478356380432,
+      "grad_norm": 0.45048559828104257,
+      "learning_rate": 3.7650272800548316e-06,
+      "loss": 0.1507,
+      "step": 1618
+    },
+    {
+      "epoch": 0.36477314332385163,
+      "grad_norm": 0.4380402419214886,
+      "learning_rate": 3.7634201282773673e-06,
+      "loss": 0.1479,
+      "step": 1619
+    },
+    {
+      "epoch": 0.36499845100966005,
+      "grad_norm": 0.4694987763481897,
+      "learning_rate": 3.7618122749697993e-06,
+      "loss": 0.1493,
+      "step": 1620
+    },
+    {
+      "epoch": 0.3652237586954685,
+      "grad_norm": 0.49229010846473126,
+      "learning_rate": 3.7602037210249077e-06,
+      "loss": 0.166,
+      "step": 1621
+    },
+    {
+      "epoch": 0.36544906638127694,
+      "grad_norm": 0.4732508852634151,
+      "learning_rate": 3.7585944673358632e-06,
+      "loss": 0.158,
+      "step": 1622
+    },
+    {
+      "epoch": 0.36567437406708536,
+      "grad_norm": 0.4922730105742576,
+      "learning_rate": 3.756984514796224e-06,
+      "loss": 0.1658,
+      "step": 1623
+    },
+    {
+      "epoch": 0.3658996817528938,
+      "grad_norm": 0.480465721750276,
+      "learning_rate": 3.7553738642999354e-06,
+      "loss": 0.1629,
+      "step": 1624
+    },
+    {
+      "epoch": 0.36612498943870225,
+      "grad_norm": 0.4555500517432691,
+      "learning_rate": 3.753762516741333e-06,
+      "loss": 0.16,
+      "step": 1625
+    },
+    {
+      "epoch": 0.36635029712451067,
+      "grad_norm": 0.46896943858709556,
+      "learning_rate": 3.7521504730151382e-06,
+      "loss": 0.1684,
+      "step": 1626
+    },
+    {
+      "epoch": 0.3665756048103191,
+      "grad_norm": 0.4746436027474732,
+      "learning_rate": 3.7505377340164585e-06,
+      "loss": 0.1634,
+      "step": 1627
+    },
+    {
+      "epoch": 0.3668009124961275,
+      "grad_norm": 0.49150528576567154,
+      "learning_rate": 3.748924300640787e-06,
+      "loss": 0.1754,
+      "step": 1628
+    },
+    {
+      "epoch": 0.367026220181936,
+      "grad_norm": 0.4800251870123433,
+      "learning_rate": 3.747310173784004e-06,
+      "loss": 0.161,
+      "step": 1629
+    },
+    {
+      "epoch": 0.3672515278677444,
+      "grad_norm": 0.4886368625564826,
+      "learning_rate": 3.745695354342374e-06,
+      "loss": 0.173,
+      "step": 1630
+    },
+    {
+      "epoch": 0.3674768355535528,
+      "grad_norm": 0.4710407314033299,
+      "learning_rate": 3.7440798432125452e-06,
+      "loss": 0.1496,
+      "step": 1631
+    },
+    {
+      "epoch": 0.36770214323936123,
+      "grad_norm": 0.44396957481507243,
+      "learning_rate": 3.742463641291552e-06,
+      "loss": 0.1489,
+      "step": 1632
+    },
+    {
+      "epoch": 0.3679274509251697,
+      "grad_norm": 0.4495509640753466,
+      "learning_rate": 3.7408467494768104e-06,
+      "loss": 0.1433,
+      "step": 1633
+    },
+    {
+      "epoch": 0.3681527586109781,
+      "grad_norm": 0.4911614375848097,
+      "learning_rate": 3.73922916866612e-06,
+      "loss": 0.1711,
+      "step": 1634
+    },
+    {
+      "epoch": 0.36837806629678654,
+      "grad_norm": 0.4925175016344445,
+      "learning_rate": 3.7376108997576628e-06,
+      "loss": 0.1625,
+      "step": 1635
+    },
+    {
+      "epoch": 0.36860337398259496,
+      "grad_norm": 0.47304790543096314,
+      "learning_rate": 3.7359919436500038e-06,
+      "loss": 0.1722,
+      "step": 1636
+    },
+    {
+      "epoch": 0.36882868166840344,
+      "grad_norm": 0.5227935062806247,
+      "learning_rate": 3.7343723012420884e-06,
+      "loss": 0.1708,
+      "step": 1637
+    },
+    {
+      "epoch": 0.36905398935421185,
+      "grad_norm": 0.43706210714839444,
+      "learning_rate": 3.7327519734332453e-06,
+      "loss": 0.1516,
+      "step": 1638
+    },
+    {
+      "epoch": 0.36927929704002027,
+      "grad_norm": 0.4483713856914514,
+      "learning_rate": 3.73113096112318e-06,
+      "loss": 0.1591,
+      "step": 1639
+    },
+    {
+      "epoch": 0.3695046047258287,
+      "grad_norm": 0.44984030973045785,
+      "learning_rate": 3.7295092652119815e-06,
+      "loss": 0.1574,
+      "step": 1640
+    },
+    {
+      "epoch": 0.36972991241163716,
+      "grad_norm": 0.45693541394796433,
+      "learning_rate": 3.7278868866001165e-06,
+      "loss": 0.161,
+      "step": 1641
+    },
+    {
+      "epoch": 0.3699552200974456,
+      "grad_norm": 0.46348952374315455,
+      "learning_rate": 3.726263826188432e-06,
+      "loss": 0.1613,
+      "step": 1642
+    },
+    {
+      "epoch": 0.370180527783254,
+      "grad_norm": 0.46604172277835615,
+      "learning_rate": 3.724640084878153e-06,
+      "loss": 0.1584,
+      "step": 1643
+    },
+    {
+      "epoch": 0.3704058354690624,
+      "grad_norm": 0.4853993400922052,
+      "learning_rate": 3.7230156635708815e-06,
+      "loss": 0.1699,
+      "step": 1644
+    },
+    {
+      "epoch": 0.3706311431548709,
+      "grad_norm": 0.441528425970984,
+      "learning_rate": 3.7213905631685988e-06,
+      "loss": 0.1547,
+      "step": 1645
+    },
+    {
+      "epoch": 0.3708564508406793,
+      "grad_norm": 0.48827412915196455,
+      "learning_rate": 3.7197647845736616e-06,
+      "loss": 0.1526,
+      "step": 1646
+    },
+    {
+      "epoch": 0.3710817585264877,
+      "grad_norm": 0.4670003645595098,
+      "learning_rate": 3.7181383286888056e-06,
+      "loss": 0.1532,
+      "step": 1647
+    },
+    {
+      "epoch": 0.37130706621229614,
+      "grad_norm": 0.46132477401273203,
+      "learning_rate": 3.7165111964171407e-06,
+      "loss": 0.156,
+      "step": 1648
+    },
+    {
+      "epoch": 0.3715323738981046,
+      "grad_norm": 0.5060266476097205,
+      "learning_rate": 3.714883388662153e-06,
+      "loss": 0.1696,
+      "step": 1649
+    },
+    {
+      "epoch": 0.37175768158391304,
+      "grad_norm": 0.4444436379093003,
+      "learning_rate": 3.7132549063277033e-06,
+      "loss": 0.1577,
+      "step": 1650
+    },
+    {
+      "epoch": 0.37198298926972145,
+      "grad_norm": 0.4985865793993255,
+      "learning_rate": 3.711625750318026e-06,
+      "loss": 0.1807,
+      "step": 1651
+    },
+    {
+      "epoch": 0.37220829695552987,
+      "grad_norm": 0.45732286320521387,
+      "learning_rate": 3.7099959215377325e-06,
+      "loss": 0.1605,
+      "step": 1652
+    },
+    {
+      "epoch": 0.37243360464133834,
+      "grad_norm": 0.4660657647459733,
+      "learning_rate": 3.7083654208918044e-06,
+      "loss": 0.1546,
+      "step": 1653
+    },
+    {
+      "epoch": 0.37265891232714676,
+      "grad_norm": 0.44304367341123163,
+      "learning_rate": 3.7067342492855997e-06,
+      "loss": 0.1521,
+      "step": 1654
+    },
+    {
+      "epoch": 0.3728842200129552,
+      "grad_norm": 0.47238935736242904,
+      "learning_rate": 3.7051024076248455e-06,
+      "loss": 0.1609,
+      "step": 1655
+    },
+    {
+      "epoch": 0.3731095276987636,
+      "grad_norm": 0.4508232710567462,
+      "learning_rate": 3.7034698968156434e-06,
+      "loss": 0.1495,
+      "step": 1656
+    },
+    {
+      "epoch": 0.3733348353845721,
+      "grad_norm": 0.5187504962903139,
+      "learning_rate": 3.7018367177644654e-06,
+      "loss": 0.1969,
+      "step": 1657
+    },
+    {
+      "epoch": 0.3735601430703805,
+      "grad_norm": 0.4859885855039429,
+      "learning_rate": 3.700202871378156e-06,
+      "loss": 0.1691,
+      "step": 1658
+    },
+    {
+      "epoch": 0.3737854507561889,
+      "grad_norm": 0.48555317612448284,
+      "learning_rate": 3.698568358563928e-06,
+      "loss": 0.1618,
+      "step": 1659
+    },
+    {
+      "epoch": 0.3740107584419973,
+      "grad_norm": 0.471500260609219,
+      "learning_rate": 3.696933180229366e-06,
+      "loss": 0.1617,
+      "step": 1660
+    },
+    {
+      "epoch": 0.3742360661278058,
+      "grad_norm": 0.5179863634705896,
+      "learning_rate": 3.6952973372824236e-06,
+      "loss": 0.1738,
+      "step": 1661
+    },
+    {
+      "epoch": 0.3744613738136142,
+      "grad_norm": 0.46447876453674614,
+      "learning_rate": 3.6936608306314227e-06,
+      "loss": 0.1505,
+      "step": 1662
+    },
+    {
+      "epoch": 0.37468668149942264,
+      "grad_norm": 0.4636206922712721,
+      "learning_rate": 3.6920236611850557e-06,
+      "loss": 0.1432,
+      "step": 1663
+    },
+    {
+      "epoch": 0.37491198918523105,
+      "grad_norm": 0.49137205139355977,
+      "learning_rate": 3.690385829852381e-06,
+      "loss": 0.1682,
+      "step": 1664
+    },
+    {
+      "epoch": 0.3751372968710395,
+      "grad_norm": 0.478847300739953,
+      "learning_rate": 3.6887473375428257e-06,
+      "loss": 0.171,
+      "step": 1665
+    },
+    {
+      "epoch": 0.37536260455684795,
+      "grad_norm": 0.5097285105087138,
+      "learning_rate": 3.6871081851661825e-06,
+      "loss": 0.172,
+      "step": 1666
+    },
+    {
+      "epoch": 0.37558791224265636,
+      "grad_norm": 0.4606989176372907,
+      "learning_rate": 3.685468373632613e-06,
+      "loss": 0.1584,
+      "step": 1667
+    },
+    {
+      "epoch": 0.37581321992846484,
+      "grad_norm": 0.47821792660078855,
+      "learning_rate": 3.6838279038526427e-06,
+      "loss": 0.1621,
+      "step": 1668
+    },
+    {
+      "epoch": 0.37603852761427325,
+      "grad_norm": 0.49790475159273734,
+      "learning_rate": 3.6821867767371634e-06,
+      "loss": 0.1671,
+      "step": 1669
+    },
+    {
+      "epoch": 0.3762638353000817,
+      "grad_norm": 0.45791610679822975,
+      "learning_rate": 3.6805449931974313e-06,
+      "loss": 0.1603,
+      "step": 1670
+    },
+    {
+      "epoch": 0.3764891429858901,
+      "grad_norm": 0.4993384427798507,
+      "learning_rate": 3.6789025541450686e-06,
+      "loss": 0.1617,
+      "step": 1671
+    },
+    {
+      "epoch": 0.37671445067169856,
+      "grad_norm": 0.4810348992142745,
+      "learning_rate": 3.67725946049206e-06,
+      "loss": 0.1598,
+      "step": 1672
+    },
+    {
+      "epoch": 0.376939758357507,
+      "grad_norm": 0.5228990391190075,
+      "learning_rate": 3.675615713150754e-06,
+      "loss": 0.1771,
+      "step": 1673
+    },
+    {
+      "epoch": 0.3771650660433154,
+      "grad_norm": 0.4933808144229279,
+      "learning_rate": 3.6739713130338617e-06,
+      "loss": 0.1766,
+      "step": 1674
+    },
+    {
+      "epoch": 0.3773903737291238,
+      "grad_norm": 0.49129051234537585,
+      "learning_rate": 3.6723262610544586e-06,
+      "loss": 0.1753,
+      "step": 1675
+    },
+    {
+      "epoch": 0.3776156814149323,
+      "grad_norm": 0.4765726245594172,
+      "learning_rate": 3.6706805581259807e-06,
+      "loss": 0.1608,
+      "step": 1676
+    },
+    {
+      "epoch": 0.3778409891007407,
+      "grad_norm": 0.48126788082063077,
+      "learning_rate": 3.669034205162224e-06,
+      "loss": 0.1618,
+      "step": 1677
+    },
+    {
+      "epoch": 0.3780662967865491,
+      "grad_norm": 0.4743792209420733,
+      "learning_rate": 3.6673872030773473e-06,
+      "loss": 0.1589,
+      "step": 1678
+    },
+    {
+      "epoch": 0.37829160447235755,
+      "grad_norm": 0.45737859410066595,
+      "learning_rate": 3.66573955278587e-06,
+      "loss": 0.1537,
+      "step": 1679
+    },
+    {
+      "epoch": 0.378516912158166,
+      "grad_norm": 0.47694611648098956,
+      "learning_rate": 3.664091255202672e-06,
+      "loss": 0.1623,
+      "step": 1680
+    },
+    {
+      "epoch": 0.37874221984397444,
+      "grad_norm": 0.4972321650519721,
+      "learning_rate": 3.662442311242989e-06,
+      "loss": 0.17,
+      "step": 1681
+    },
+    {
+      "epoch": 0.37896752752978285,
+      "grad_norm": 0.49956960858663274,
+      "learning_rate": 3.66079272182242e-06,
+      "loss": 0.1743,
+      "step": 1682
+    },
+    {
+      "epoch": 0.3791928352155913,
+      "grad_norm": 0.4628531179808592,
+      "learning_rate": 3.6591424878569203e-06,
+      "loss": 0.1615,
+      "step": 1683
+    },
+    {
+      "epoch": 0.37941814290139975,
+      "grad_norm": 0.506047314167328,
+      "learning_rate": 3.657491610262802e-06,
+      "loss": 0.1785,
+      "step": 1684
+    },
+    {
+      "epoch": 0.37964345058720816,
+      "grad_norm": 0.4616821462609864,
+      "learning_rate": 3.655840089956738e-06,
+      "loss": 0.1603,
+      "step": 1685
+    },
+    {
+      "epoch": 0.3798687582730166,
+      "grad_norm": 0.44081782340460696,
+      "learning_rate": 3.654187927855754e-06,
+      "loss": 0.156,
+      "step": 1686
+    },
+    {
+      "epoch": 0.380094065958825,
+      "grad_norm": 0.465733362020006,
+      "learning_rate": 3.6525351248772357e-06,
+      "loss": 0.1438,
+      "step": 1687
+    },
+    {
+      "epoch": 0.3803193736446335,
+      "grad_norm": 0.49319990988355183,
+      "learning_rate": 3.6508816819389216e-06,
+      "loss": 0.1655,
+      "step": 1688
+    },
+    {
+      "epoch": 0.3805446813304419,
+      "grad_norm": 0.47847792905995146,
+      "learning_rate": 3.6492275999589065e-06,
+      "loss": 0.1537,
+      "step": 1689
+    },
+    {
+      "epoch": 0.3807699890162503,
+      "grad_norm": 0.4929434149921711,
+      "learning_rate": 3.6475728798556426e-06,
+      "loss": 0.1675,
+      "step": 1690
+    },
+    {
+      "epoch": 0.38099529670205873,
+      "grad_norm": 0.48998776304248226,
+      "learning_rate": 3.645917522547933e-06,
+      "loss": 0.1749,
+      "step": 1691
+    },
+    {
+      "epoch": 0.3812206043878672,
+      "grad_norm": 0.4906889531338348,
+      "learning_rate": 3.6442615289549354e-06,
+      "loss": 0.1623,
+      "step": 1692
+    },
+    {
+      "epoch": 0.3814459120736756,
+      "grad_norm": 0.5004684870879217,
+      "learning_rate": 3.6426048999961626e-06,
+      "loss": 0.171,
+      "step": 1693
+    },
+    {
+      "epoch": 0.38167121975948404,
+      "grad_norm": 0.4722242162068364,
+      "learning_rate": 3.6409476365914786e-06,
+      "loss": 0.1591,
+      "step": 1694
+    },
+    {
+      "epoch": 0.38189652744529246,
+      "grad_norm": 0.5011760714533439,
+      "learning_rate": 3.6392897396610992e-06,
+      "loss": 0.1753,
+      "step": 1695
+    },
+    {
+      "epoch": 0.38212183513110093,
+      "grad_norm": 0.47823957791904037,
+      "learning_rate": 3.6376312101255934e-06,
+      "loss": 0.1552,
+      "step": 1696
+    },
+    {
+      "epoch": 0.38234714281690935,
+      "grad_norm": 0.48094097708510114,
+      "learning_rate": 3.6359720489058804e-06,
+      "loss": 0.1586,
+      "step": 1697
+    },
+    {
+      "epoch": 0.38257245050271776,
+      "grad_norm": 0.485064759706226,
+      "learning_rate": 3.6343122569232313e-06,
+      "loss": 0.1708,
+      "step": 1698
+    },
+    {
+      "epoch": 0.3827977581885262,
+      "grad_norm": 0.4724466085894854,
+      "learning_rate": 3.6326518350992657e-06,
+      "loss": 0.149,
+      "step": 1699
+    },
+    {
+      "epoch": 0.38302306587433466,
+      "grad_norm": 0.48086104377028743,
+      "learning_rate": 3.6309907843559542e-06,
+      "loss": 0.1575,
+      "step": 1700
+    },
+    {
+      "epoch": 0.3832483735601431,
+      "grad_norm": 0.4589335602870298,
+      "learning_rate": 3.6293291056156178e-06,
+      "loss": 0.1553,
+      "step": 1701
+    },
+    {
+      "epoch": 0.3834736812459515,
+      "grad_norm": 0.49392644134888775,
+      "learning_rate": 3.6276667998009242e-06,
+      "loss": 0.1605,
+      "step": 1702
+    },
+    {
+      "epoch": 0.3836989889317599,
+      "grad_norm": 0.49623607357562033,
+      "learning_rate": 3.626003867834888e-06,
+      "loss": 0.1715,
+      "step": 1703
+    },
+    {
+      "epoch": 0.3839242966175684,
+      "grad_norm": 0.5028003735238179,
+      "learning_rate": 3.624340310640875e-06,
+      "loss": 0.1715,
+      "step": 1704
+    },
+    {
+      "epoch": 0.3841496043033768,
+      "grad_norm": 0.5246279321726527,
+      "learning_rate": 3.6226761291425956e-06,
+      "loss": 0.1677,
+      "step": 1705
+    },
+    {
+      "epoch": 0.3843749119891852,
+      "grad_norm": 0.4714051843773443,
+      "learning_rate": 3.621011324264109e-06,
+      "loss": 0.1664,
+      "step": 1706
+    },
+    {
+      "epoch": 0.38460021967499364,
+      "grad_norm": 0.47138163904110986,
+      "learning_rate": 3.6193458969298184e-06,
+      "loss": 0.1764,
+      "step": 1707
+    },
+    {
+      "epoch": 0.3848255273608021,
+      "grad_norm": 0.4923417840544699,
+      "learning_rate": 3.617679848064474e-06,
+      "loss": 0.1597,
+      "step": 1708
+    },
+    {
+      "epoch": 0.38505083504661053,
+      "grad_norm": 0.4649501594809374,
+      "learning_rate": 3.6160131785931695e-06,
+      "loss": 0.158,
+      "step": 1709
+    },
+    {
+      "epoch": 0.38527614273241895,
+      "grad_norm": 0.4682784221008982,
+      "learning_rate": 3.6143458894413463e-06,
+      "loss": 0.1598,
+      "step": 1710
+    },
+    {
+      "epoch": 0.38550145041822736,
+      "grad_norm": 0.4559827677564158,
+      "learning_rate": 3.6126779815347863e-06,
+      "loss": 0.1559,
+      "step": 1711
+    },
+    {
+      "epoch": 0.38572675810403584,
+      "grad_norm": 0.4422578569156396,
+      "learning_rate": 3.611009455799617e-06,
+      "loss": 0.1509,
+      "step": 1712
+    },
+    {
+      "epoch": 0.38595206578984426,
+      "grad_norm": 0.4437396743103725,
+      "learning_rate": 3.609340313162309e-06,
+      "loss": 0.147,
+      "step": 1713
+    },
+    {
+      "epoch": 0.3861773734756527,
+      "grad_norm": 0.4720363562053173,
+      "learning_rate": 3.6076705545496743e-06,
+      "loss": 0.1727,
+      "step": 1714
+    },
+    {
+      "epoch": 0.38640268116146115,
+      "grad_norm": 0.4443211995923577,
+      "learning_rate": 3.606000180888868e-06,
+      "loss": 0.1467,
+      "step": 1715
+    },
+    {
+      "epoch": 0.38662798884726957,
+      "grad_norm": 0.47935065592719434,
+      "learning_rate": 3.604329193107386e-06,
+      "loss": 0.1543,
+      "step": 1716
+    },
+    {
+      "epoch": 0.386853296533078,
+      "grad_norm": 0.5092502507366087,
+      "learning_rate": 3.6026575921330665e-06,
+      "loss": 0.17,
+      "step": 1717
+    },
+    {
+      "epoch": 0.3870786042188864,
+      "grad_norm": 0.48204778905856294,
+      "learning_rate": 3.600985378894086e-06,
+      "loss": 0.1628,
+      "step": 1718
+    },
+    {
+      "epoch": 0.3873039119046949,
+      "grad_norm": 0.48203224699763203,
+      "learning_rate": 3.5993125543189634e-06,
+      "loss": 0.1676,
+      "step": 1719
+    },
+    {
+      "epoch": 0.3875292195905033,
+      "grad_norm": 0.454587963068617,
+      "learning_rate": 3.5976391193365544e-06,
+      "loss": 0.1509,
+      "step": 1720
+    },
+    {
+      "epoch": 0.3877545272763117,
+      "grad_norm": 0.498846730629908,
+      "learning_rate": 3.5959650748760562e-06,
+      "loss": 0.1753,
+      "step": 1721
+    },
+    {
+      "epoch": 0.38797983496212013,
+      "grad_norm": 0.45540583194092865,
+      "learning_rate": 3.5942904218670025e-06,
+      "loss": 0.1609,
+      "step": 1722
+    },
+    {
+      "epoch": 0.3882051426479286,
+      "grad_norm": 0.4756057121487946,
+      "learning_rate": 3.592615161239267e-06,
+      "loss": 0.1636,
+      "step": 1723
+    },
+    {
+      "epoch": 0.388430450333737,
+      "grad_norm": 0.47297457111173763,
+      "learning_rate": 3.590939293923058e-06,
+      "loss": 0.1701,
+      "step": 1724
+    },
+    {
+      "epoch": 0.38865575801954544,
+      "grad_norm": 0.5041510529405214,
+      "learning_rate": 3.5892628208489226e-06,
+      "loss": 0.1587,
+      "step": 1725
+    },
+    {
+      "epoch": 0.38888106570535386,
+      "grad_norm": 0.46158951154891076,
+      "learning_rate": 3.5875857429477447e-06,
+      "loss": 0.1594,
+      "step": 1726
+    },
+    {
+      "epoch": 0.38910637339116233,
+      "grad_norm": 0.4415370624886341,
+      "learning_rate": 3.585908061150741e-06,
+      "loss": 0.1493,
+      "step": 1727
+    },
+    {
+      "epoch": 0.38933168107697075,
+      "grad_norm": 0.4744067839485117,
+      "learning_rate": 3.584229776389468e-06,
+      "loss": 0.1685,
+      "step": 1728
+    },
+    {
+      "epoch": 0.38955698876277917,
+      "grad_norm": 0.5160028615095501,
+      "learning_rate": 3.5825508895958143e-06,
+      "loss": 0.1733,
+      "step": 1729
+    },
+    {
+      "epoch": 0.3897822964485876,
+      "grad_norm": 0.5072082547164747,
+      "learning_rate": 3.580871401702002e-06,
+      "loss": 0.1717,
+      "step": 1730
+    },
+    {
+      "epoch": 0.39000760413439606,
+      "grad_norm": 0.44889530087849144,
+      "learning_rate": 3.5791913136405883e-06,
+      "loss": 0.1512,
+      "step": 1731
+    },
+    {
+      "epoch": 0.3902329118202045,
+      "grad_norm": 0.47243671574782475,
+      "learning_rate": 3.5775106263444644e-06,
+      "loss": 0.1571,
+      "step": 1732
+    },
+    {
+      "epoch": 0.3904582195060129,
+      "grad_norm": 0.4673534957109278,
+      "learning_rate": 3.5758293407468525e-06,
+      "loss": 0.1536,
+      "step": 1733
+    },
+    {
+      "epoch": 0.3906835271918213,
+      "grad_norm": 0.5182709051377471,
+      "learning_rate": 3.5741474577813086e-06,
+      "loss": 0.1725,
+      "step": 1734
+    },
+    {
+      "epoch": 0.3909088348776298,
+      "grad_norm": 0.5045935815654671,
+      "learning_rate": 3.572464978381719e-06,
+      "loss": 0.17,
+      "step": 1735
+    },
+    {
+      "epoch": 0.3911341425634382,
+      "grad_norm": 0.4855265240200406,
+      "learning_rate": 3.570781903482302e-06,
+      "loss": 0.174,
+      "step": 1736
+    },
+    {
+      "epoch": 0.3913594502492466,
+      "grad_norm": 0.45358032477406396,
+      "learning_rate": 3.569098234017606e-06,
+      "loss": 0.1518,
+      "step": 1737
+    },
+    {
+      "epoch": 0.39158475793505504,
+      "grad_norm": 0.5160232778445221,
+      "learning_rate": 3.5674139709225104e-06,
+      "loss": 0.1684,
+      "step": 1738
+    },
+    {
+      "epoch": 0.3918100656208635,
+      "grad_norm": 0.4803803541458198,
+      "learning_rate": 3.565729115132224e-06,
+      "loss": 0.1662,
+      "step": 1739
+    },
+    {
+      "epoch": 0.39203537330667193,
+      "grad_norm": 0.4787719969692587,
+      "learning_rate": 3.5640436675822833e-06,
+      "loss": 0.1565,
+      "step": 1740
+    },
+    {
+      "epoch": 0.39226068099248035,
+      "grad_norm": 0.48099904622308787,
+      "learning_rate": 3.5623576292085555e-06,
+      "loss": 0.1659,
+      "step": 1741
+    },
+    {
+      "epoch": 0.39248598867828877,
+      "grad_norm": 0.48566284958274347,
+      "learning_rate": 3.5606710009472335e-06,
+      "loss": 0.1562,
+      "step": 1742
+    },
+    {
+      "epoch": 0.39271129636409724,
+      "grad_norm": 0.502006974708806,
+      "learning_rate": 3.558983783734841e-06,
+      "loss": 0.1572,
+      "step": 1743
+    },
+    {
+      "epoch": 0.39293660404990566,
+      "grad_norm": 0.4542809843689211,
+      "learning_rate": 3.5572959785082264e-06,
+      "loss": 0.1519,
+      "step": 1744
+    },
+    {
+      "epoch": 0.3931619117357141,
+      "grad_norm": 0.4747827690867886,
+      "learning_rate": 3.5556075862045636e-06,
+      "loss": 0.1591,
+      "step": 1745
+    },
+    {
+      "epoch": 0.3933872194215225,
+      "grad_norm": 0.4862382157467648,
+      "learning_rate": 3.5539186077613562e-06,
+      "loss": 0.1517,
+      "step": 1746
+    },
+    {
+      "epoch": 0.39361252710733097,
+      "grad_norm": 0.48020012626984315,
+      "learning_rate": 3.552229044116428e-06,
+      "loss": 0.1682,
+      "step": 1747
+    },
+    {
+      "epoch": 0.3938378347931394,
+      "grad_norm": 0.4669769781915097,
+      "learning_rate": 3.5505388962079337e-06,
+      "loss": 0.1565,
+      "step": 1748
+    },
+    {
+      "epoch": 0.3940631424789478,
+      "grad_norm": 0.4648212358293205,
+      "learning_rate": 3.548848164974347e-06,
+      "loss": 0.1523,
+      "step": 1749
+    },
+    {
+      "epoch": 0.3942884501647562,
+      "grad_norm": 0.4987953329315867,
+      "learning_rate": 3.54715685135447e-06,
+      "loss": 0.1847,
+      "step": 1750
+    },
+    {
+      "epoch": 0.3945137578505647,
+      "grad_norm": 0.44830900353615505,
+      "learning_rate": 3.545464956287425e-06,
+      "loss": 0.1544,
+      "step": 1751
+    },
+    {
+      "epoch": 0.3947390655363731,
+      "grad_norm": 0.44263663083561267,
+      "learning_rate": 3.5437724807126583e-06,
+      "loss": 0.1486,
+      "step": 1752
+    },
+    {
+      "epoch": 0.39496437322218153,
+      "grad_norm": 0.46596869817108705,
+      "learning_rate": 3.542079425569938e-06,
+      "loss": 0.1591,
+      "step": 1753
+    },
+    {
+      "epoch": 0.39518968090798995,
+      "grad_norm": 0.4619508102805029,
+      "learning_rate": 3.5403857917993554e-06,
+      "loss": 0.1615,
+      "step": 1754
+    },
+    {
+      "epoch": 0.3954149885937984,
+      "grad_norm": 0.5113106262452023,
+      "learning_rate": 3.5386915803413234e-06,
+      "loss": 0.1746,
+      "step": 1755
+    },
+    {
+      "epoch": 0.39564029627960684,
+      "grad_norm": 0.47742352068847194,
+      "learning_rate": 3.5369967921365718e-06,
+      "loss": 0.1566,
+      "step": 1756
+    },
+    {
+      "epoch": 0.39586560396541526,
+      "grad_norm": 0.48642828981283603,
+      "learning_rate": 3.5353014281261545e-06,
+      "loss": 0.1565,
+      "step": 1757
+    },
+    {
+      "epoch": 0.3960909116512237,
+      "grad_norm": 0.48189096819477817,
+      "learning_rate": 3.5336054892514437e-06,
+      "loss": 0.1691,
+      "step": 1758
+    },
+    {
+      "epoch": 0.39631621933703215,
+      "grad_norm": 0.48477286061008706,
+      "learning_rate": 3.531908976454132e-06,
+      "loss": 0.1669,
+      "step": 1759
+    },
+    {
+      "epoch": 0.39654152702284057,
+      "grad_norm": 0.44895590316474676,
+      "learning_rate": 3.530211890676229e-06,
+      "loss": 0.1443,
+      "step": 1760
+    },
+    {
+      "epoch": 0.396766834708649,
+      "grad_norm": 0.46011556543301124,
+      "learning_rate": 3.528514232860063e-06,
+      "loss": 0.167,
+      "step": 1761
+    },
+    {
+      "epoch": 0.3969921423944574,
+      "grad_norm": 0.4858974568703489,
+      "learning_rate": 3.52681600394828e-06,
+      "loss": 0.1719,
+      "step": 1762
+    },
+    {
+      "epoch": 0.3972174500802659,
+      "grad_norm": 0.46589057661030264,
+      "learning_rate": 3.525117204883844e-06,
+      "loss": 0.1569,
+      "step": 1763
+    },
+    {
+      "epoch": 0.3974427577660743,
+      "grad_norm": 0.4945996442906626,
+      "learning_rate": 3.5234178366100343e-06,
+      "loss": 0.1698,
+      "step": 1764
+    },
+    {
+      "epoch": 0.3976680654518827,
+      "grad_norm": 0.5230375509533259,
+      "learning_rate": 3.5217179000704467e-06,
+      "loss": 0.1529,
+      "step": 1765
+    },
+    {
+      "epoch": 0.3978933731376912,
+      "grad_norm": 0.4744484175293173,
+      "learning_rate": 3.520017396208993e-06,
+      "loss": 0.1653,
+      "step": 1766
+    },
+    {
+      "epoch": 0.3981186808234996,
+      "grad_norm": 0.4897183630580415,
+      "learning_rate": 3.518316325969899e-06,
+      "loss": 0.1691,
+      "step": 1767
+    },
+    {
+      "epoch": 0.398343988509308,
+      "grad_norm": 0.501512935731379,
+      "learning_rate": 3.5166146902977055e-06,
+      "loss": 0.1732,
+      "step": 1768
+    },
+    {
+      "epoch": 0.39856929619511644,
+      "grad_norm": 0.4910117121160473,
+      "learning_rate": 3.514912490137268e-06,
+      "loss": 0.1575,
+      "step": 1769
+    },
+    {
+      "epoch": 0.3987946038809249,
+      "grad_norm": 0.5124656318430874,
+      "learning_rate": 3.5132097264337546e-06,
+      "loss": 0.1647,
+      "step": 1770
+    },
+    {
+      "epoch": 0.39901991156673333,
+      "grad_norm": 0.4908288371907048,
+      "learning_rate": 3.5115064001326467e-06,
+      "loss": 0.1649,
+      "step": 1771
+    },
+    {
+      "epoch": 0.39924521925254175,
+      "grad_norm": 0.48086723360755884,
+      "learning_rate": 3.5098025121797375e-06,
+      "loss": 0.1626,
+      "step": 1772
+    },
+    {
+      "epoch": 0.39947052693835017,
+      "grad_norm": 0.4692892966024187,
+      "learning_rate": 3.508098063521134e-06,
+      "loss": 0.1591,
+      "step": 1773
+    },
+    {
+      "epoch": 0.39969583462415864,
+      "grad_norm": 0.5065104083880511,
+      "learning_rate": 3.5063930551032494e-06,
+      "loss": 0.1653,
+      "step": 1774
+    },
+    {
+      "epoch": 0.39992114230996706,
+      "grad_norm": 0.47844515863928405,
+      "learning_rate": 3.504687487872815e-06,
+      "loss": 0.159,
+      "step": 1775
+    },
+    {
+      "epoch": 0.4001464499957755,
+      "grad_norm": 0.5296198292166024,
+      "learning_rate": 3.5029813627768665e-06,
+      "loss": 0.1796,
+      "step": 1776
+    },
+    {
+      "epoch": 0.4003717576815839,
+      "grad_norm": 0.48975396814814165,
+      "learning_rate": 3.501274680762753e-06,
+      "loss": 0.1517,
+      "step": 1777
+    },
+    {
+      "epoch": 0.40059706536739237,
+      "grad_norm": 0.47662762225410193,
+      "learning_rate": 3.499567442778131e-06,
+      "loss": 0.1602,
+      "step": 1778
+    },
+    {
+      "epoch": 0.4008223730532008,
+      "grad_norm": 0.49921176967190556,
+      "learning_rate": 3.497859649770965e-06,
+      "loss": 0.1666,
+      "step": 1779
+    },
+    {
+      "epoch": 0.4010476807390092,
+      "grad_norm": 0.4678569612081498,
+      "learning_rate": 3.49615130268953e-06,
+      "loss": 0.1548,
+      "step": 1780
+    },
+    {
+      "epoch": 0.4012729884248176,
+      "grad_norm": 0.492057854759809,
+      "learning_rate": 3.494442402482407e-06,
+      "loss": 0.1685,
+      "step": 1781
+    },
+    {
+      "epoch": 0.4014982961106261,
+      "grad_norm": 0.4657679154537899,
+      "learning_rate": 3.4927329500984857e-06,
+      "loss": 0.1527,
+      "step": 1782
+    },
+    {
+      "epoch": 0.4017236037964345,
+      "grad_norm": 0.49090571162135244,
+      "learning_rate": 3.4910229464869594e-06,
+      "loss": 0.1668,
+      "step": 1783
+    },
+    {
+      "epoch": 0.40194891148224293,
+      "grad_norm": 0.4774053398472881,
+      "learning_rate": 3.489312392597331e-06,
+      "loss": 0.1609,
+      "step": 1784
+    },
+    {
+      "epoch": 0.40217421916805135,
+      "grad_norm": 0.5139863051826496,
+      "learning_rate": 3.4876012893794053e-06,
+      "loss": 0.184,
+      "step": 1785
+    },
+    {
+      "epoch": 0.4023995268538598,
+      "grad_norm": 0.44779930986429073,
+      "learning_rate": 3.4858896377832966e-06,
+      "loss": 0.1407,
+      "step": 1786
+    },
+    {
+      "epoch": 0.40262483453966824,
+      "grad_norm": 0.49744330625774086,
+      "learning_rate": 3.4841774387594202e-06,
+      "loss": 0.1739,
+      "step": 1787
+    },
+    {
+      "epoch": 0.40285014222547666,
+      "grad_norm": 0.43623504745923614,
+      "learning_rate": 3.482464693258496e-06,
+      "loss": 0.1453,
+      "step": 1788
+    },
+    {
+      "epoch": 0.4030754499112851,
+      "grad_norm": 0.4794462886411687,
+      "learning_rate": 3.4807514022315473e-06,
+      "loss": 0.1564,
+      "step": 1789
+    },
+    {
+      "epoch": 0.40330075759709355,
+      "grad_norm": 0.4983964466277479,
+      "learning_rate": 3.4790375666299026e-06,
+      "loss": 0.1555,
+      "step": 1790
+    },
+    {
+      "epoch": 0.40352606528290197,
+      "grad_norm": 0.4818184175327813,
+      "learning_rate": 3.4773231874051893e-06,
+      "loss": 0.1661,
+      "step": 1791
+    },
+    {
+      "epoch": 0.4037513729687104,
+      "grad_norm": 0.4833608794554769,
+      "learning_rate": 3.4756082655093387e-06,
+      "loss": 0.1651,
+      "step": 1792
+    },
+    {
+      "epoch": 0.4039766806545188,
+      "grad_norm": 0.4771175845930751,
+      "learning_rate": 3.4738928018945828e-06,
+      "loss": 0.1584,
+      "step": 1793
+    },
+    {
+      "epoch": 0.4042019883403273,
+      "grad_norm": 0.4471819840046623,
+      "learning_rate": 3.4721767975134557e-06,
+      "loss": 0.1504,
+      "step": 1794
+    },
+    {
+      "epoch": 0.4044272960261357,
+      "grad_norm": 0.47232101345411487,
+      "learning_rate": 3.470460253318789e-06,
+      "loss": 0.1608,
+      "step": 1795
+    },
+    {
+      "epoch": 0.4046526037119441,
+      "grad_norm": 0.4797922242264535,
+      "learning_rate": 3.4687431702637165e-06,
+      "loss": 0.1675,
+      "step": 1796
+    },
+    {
+      "epoch": 0.40487791139775253,
+      "grad_norm": 0.49561782968484025,
+      "learning_rate": 3.4670255493016715e-06,
+      "loss": 0.1718,
+      "step": 1797
+    },
+    {
+      "epoch": 0.405103219083561,
+      "grad_norm": 0.4584936281742346,
+      "learning_rate": 3.465307391386383e-06,
+      "loss": 0.1442,
+      "step": 1798
+    },
+    {
+      "epoch": 0.4053285267693694,
+      "grad_norm": 0.47911010769893764,
+      "learning_rate": 3.4635886974718814e-06,
+      "loss": 0.1696,
+      "step": 1799
+    },
+    {
+      "epoch": 0.40555383445517784,
+      "grad_norm": 0.4559976548205687,
+      "learning_rate": 3.4618694685124927e-06,
+      "loss": 0.1518,
+      "step": 1800
+    },
+    {
+      "epoch": 0.40577914214098626,
+      "grad_norm": 0.48899872047373,
+      "learning_rate": 3.4601497054628407e-06,
+      "loss": 0.172,
+      "step": 1801
+    },
+    {
+      "epoch": 0.40600444982679473,
+      "grad_norm": 0.4644054553830585,
+      "learning_rate": 3.458429409277846e-06,
+      "loss": 0.1579,
+      "step": 1802
+    },
+    {
+      "epoch": 0.40622975751260315,
+      "grad_norm": 0.4463344674611823,
+      "learning_rate": 3.4567085809127247e-06,
+      "loss": 0.1563,
+      "step": 1803
+    },
+    {
+      "epoch": 0.40645506519841157,
+      "grad_norm": 0.4719464396077941,
+      "learning_rate": 3.454987221322989e-06,
+      "loss": 0.1675,
+      "step": 1804
+    },
+    {
+      "epoch": 0.40668037288422,
+      "grad_norm": 0.46911138126559293,
+      "learning_rate": 3.4532653314644453e-06,
+      "loss": 0.1608,
+      "step": 1805
+    },
+    {
+      "epoch": 0.40690568057002846,
+      "grad_norm": 0.46701650488610563,
+      "learning_rate": 3.4515429122931955e-06,
+      "loss": 0.1562,
+      "step": 1806
+    },
+    {
+      "epoch": 0.4071309882558369,
+      "grad_norm": 0.4345692567585763,
+      "learning_rate": 3.4498199647656335e-06,
+      "loss": 0.1489,
+      "step": 1807
+    },
+    {
+      "epoch": 0.4073562959416453,
+      "grad_norm": 0.4732432278089056,
+      "learning_rate": 3.4480964898384495e-06,
+      "loss": 0.1566,
+      "step": 1808
+    },
+    {
+      "epoch": 0.4075816036274537,
+      "grad_norm": 0.49232155873266253,
+      "learning_rate": 3.4463724884686234e-06,
+      "loss": 0.1733,
+      "step": 1809
+    },
+    {
+      "epoch": 0.4078069113132622,
+      "grad_norm": 0.47860152710599685,
+      "learning_rate": 3.44464796161343e-06,
+      "loss": 0.1583,
+      "step": 1810
+    },
+    {
+      "epoch": 0.4080322189990706,
+      "grad_norm": 0.5036762938508624,
+      "learning_rate": 3.4429229102304336e-06,
+      "loss": 0.174,
+      "step": 1811
+    },
+    {
+      "epoch": 0.408257526684879,
+      "grad_norm": 0.5282822426431432,
+      "learning_rate": 3.4411973352774917e-06,
+      "loss": 0.1734,
+      "step": 1812
+    },
+    {
+      "epoch": 0.4084828343706875,
+      "grad_norm": 0.4445382895295483,
+      "learning_rate": 3.4394712377127524e-06,
+      "loss": 0.1593,
+      "step": 1813
+    },
+    {
+      "epoch": 0.4087081420564959,
+      "grad_norm": 0.46841840331802137,
+      "learning_rate": 3.437744618494653e-06,
+      "loss": 0.161,
+      "step": 1814
+    },
+    {
+      "epoch": 0.40893344974230433,
+      "grad_norm": 0.490508242367761,
+      "learning_rate": 3.4360174785819196e-06,
+      "loss": 0.1559,
+      "step": 1815
+    },
+    {
+      "epoch": 0.40915875742811275,
+      "grad_norm": 0.48426695631141686,
+      "learning_rate": 3.4342898189335692e-06,
+      "loss": 0.1583,
+      "step": 1816
+    },
+    {
+      "epoch": 0.4093840651139212,
+      "grad_norm": 0.4367959518819385,
+      "learning_rate": 3.432561640508908e-06,
+      "loss": 0.1462,
+      "step": 1817
+    },
+    {
+      "epoch": 0.40960937279972964,
+      "grad_norm": 0.4700483688647926,
+      "learning_rate": 3.4308329442675276e-06,
+      "loss": 0.1551,
+      "step": 1818
+    },
+    {
+      "epoch": 0.40983468048553806,
+      "grad_norm": 0.435178736238957,
+      "learning_rate": 3.4291037311693088e-06,
+      "loss": 0.145,
+      "step": 1819
+    },
+    {
+      "epoch": 0.4100599881713465,
+      "grad_norm": 0.4887752868821639,
+      "learning_rate": 3.42737400217442e-06,
+      "loss": 0.1619,
+      "step": 1820
+    },
+    {
+      "epoch": 0.41028529585715495,
+      "grad_norm": 0.4538849148397559,
+      "learning_rate": 3.4256437582433144e-06,
+      "loss": 0.1531,
+      "step": 1821
+    },
+    {
+      "epoch": 0.41051060354296337,
+      "grad_norm": 0.46154436604669785,
+      "learning_rate": 3.423913000336732e-06,
+      "loss": 0.1638,
+      "step": 1822
+    },
+    {
+      "epoch": 0.4107359112287718,
+      "grad_norm": 0.43127529204771314,
+      "learning_rate": 3.422181729415699e-06,
+      "loss": 0.1422,
+      "step": 1823
+    },
+    {
+      "epoch": 0.4109612189145802,
+      "grad_norm": 0.5198286942990128,
+      "learning_rate": 3.4204499464415253e-06,
+      "loss": 0.1668,
+      "step": 1824
+    },
+    {
+      "epoch": 0.4111865266003887,
+      "grad_norm": 0.4687407439728831,
+      "learning_rate": 3.418717652375805e-06,
+      "loss": 0.1438,
+      "step": 1825
+    },
+    {
+      "epoch": 0.4114118342861971,
+      "grad_norm": 0.4821244881625982,
+      "learning_rate": 3.4169848481804165e-06,
+      "loss": 0.1688,
+      "step": 1826
+    },
+    {
+      "epoch": 0.4116371419720055,
+      "grad_norm": 0.46763244055600944,
+      "learning_rate": 3.415251534817521e-06,
+      "loss": 0.1612,
+      "step": 1827
+    },
+    {
+      "epoch": 0.41186244965781393,
+      "grad_norm": 0.48728207989142946,
+      "learning_rate": 3.4135177132495632e-06,
+      "loss": 0.1624,
+      "step": 1828
+    },
+    {
+      "epoch": 0.4120877573436224,
+      "grad_norm": 0.5312486959140987,
+      "learning_rate": 3.4117833844392704e-06,
+      "loss": 0.1735,
+      "step": 1829
+    },
+    {
+      "epoch": 0.4123130650294308,
+      "grad_norm": 0.4692598649036747,
+      "learning_rate": 3.41004854934965e-06,
+      "loss": 0.1542,
+      "step": 1830
+    },
+    {
+      "epoch": 0.41253837271523924,
+      "grad_norm": 0.46784467253864104,
+      "learning_rate": 3.4083132089439912e-06,
+      "loss": 0.1685,
+      "step": 1831
+    },
+    {
+      "epoch": 0.41276368040104766,
+      "grad_norm": 0.4670753302202224,
+      "learning_rate": 3.406577364185864e-06,
+      "loss": 0.1642,
+      "step": 1832
+    },
+    {
+      "epoch": 0.41298898808685613,
+      "grad_norm": 0.49989102638595734,
+      "learning_rate": 3.404841016039118e-06,
+      "loss": 0.1763,
+      "step": 1833
+    },
+    {
+      "epoch": 0.41321429577266455,
+      "grad_norm": 0.49099336796274373,
+      "learning_rate": 3.403104165467883e-06,
+      "loss": 0.1707,
+      "step": 1834
+    },
+    {
+      "epoch": 0.41343960345847297,
+      "grad_norm": 0.48820792687632875,
+      "learning_rate": 3.4013668134365675e-06,
+      "loss": 0.1537,
+      "step": 1835
+    },
+    {
+      "epoch": 0.4136649111442814,
+      "grad_norm": 0.47863203313919633,
+      "learning_rate": 3.399628960909857e-06,
+      "loss": 0.1609,
+      "step": 1836
+    },
+    {
+      "epoch": 0.41389021883008986,
+      "grad_norm": 0.4881564151016199,
+      "learning_rate": 3.397890608852718e-06,
+      "loss": 0.1651,
+      "step": 1837
+    },
+    {
+      "epoch": 0.4141155265158983,
+      "grad_norm": 0.47544325425792294,
+      "learning_rate": 3.3961517582303916e-06,
+      "loss": 0.1637,
+      "step": 1838
+    },
+    {
+      "epoch": 0.4143408342017067,
+      "grad_norm": 0.49922802100215147,
+      "learning_rate": 3.394412410008397e-06,
+      "loss": 0.1647,
+      "step": 1839
+    },
+    {
+      "epoch": 0.4145661418875151,
+      "grad_norm": 0.4913684385302216,
+      "learning_rate": 3.39267256515253e-06,
+      "loss": 0.1641,
+      "step": 1840
+    },
+    {
+      "epoch": 0.4147914495733236,
+      "grad_norm": 0.4471757683297294,
+      "learning_rate": 3.3909322246288606e-06,
+      "loss": 0.141,
+      "step": 1841
+    },
+    {
+      "epoch": 0.415016757259132,
+      "grad_norm": 0.48812530948929644,
+      "learning_rate": 3.3891913894037354e-06,
+      "loss": 0.1645,
+      "step": 1842
+    },
+    {
+      "epoch": 0.4152420649449404,
+      "grad_norm": 0.5119344738587837,
+      "learning_rate": 3.3874500604437752e-06,
+      "loss": 0.1548,
+      "step": 1843
+    },
+    {
+      "epoch": 0.41546737263074884,
+      "grad_norm": 0.4946995583219049,
+      "learning_rate": 3.385708238715876e-06,
+      "loss": 0.1539,
+      "step": 1844
+    },
+    {
+      "epoch": 0.4156926803165573,
+      "grad_norm": 0.44974296921309803,
+      "learning_rate": 3.3839659251872054e-06,
+      "loss": 0.1462,
+      "step": 1845
+    },
+    {
+      "epoch": 0.41591798800236573,
+      "grad_norm": 0.5062284794310744,
+      "learning_rate": 3.3822231208252053e-06,
+      "loss": 0.1617,
+      "step": 1846
+    },
+    {
+      "epoch": 0.41614329568817415,
+      "grad_norm": 0.5517545583347794,
+      "learning_rate": 3.38047982659759e-06,
+      "loss": 0.1753,
+      "step": 1847
+    },
+    {
+      "epoch": 0.41636860337398257,
+      "grad_norm": 0.5085439188834923,
+      "learning_rate": 3.3787360434723466e-06,
+      "loss": 0.1629,
+      "step": 1848
+    },
+    {
+      "epoch": 0.41659391105979104,
+      "grad_norm": 0.5002540243440992,
+      "learning_rate": 3.3769917724177315e-06,
+      "loss": 0.1725,
+      "step": 1849
+    },
+    {
+      "epoch": 0.41681921874559946,
+      "grad_norm": 0.4825881609302707,
+      "learning_rate": 3.3752470144022745e-06,
+      "loss": 0.1574,
+      "step": 1850
+    },
+    {
+      "epoch": 0.4170445264314079,
+      "grad_norm": 0.5424215184923034,
+      "learning_rate": 3.3735017703947748e-06,
+      "loss": 0.1742,
+      "step": 1851
+    },
+    {
+      "epoch": 0.4172698341172163,
+      "grad_norm": 0.5082325374018036,
+      "learning_rate": 3.371756041364301e-06,
+      "loss": 0.154,
+      "step": 1852
+    },
+    {
+      "epoch": 0.41749514180302477,
+      "grad_norm": 0.4676831647620805,
+      "learning_rate": 3.370009828280191e-06,
+      "loss": 0.1527,
+      "step": 1853
+    },
+    {
+      "epoch": 0.4177204494888332,
+      "grad_norm": 0.46679801722571546,
+      "learning_rate": 3.3682631321120507e-06,
+      "loss": 0.1698,
+      "step": 1854
+    },
+    {
+      "epoch": 0.4179457571746416,
+      "grad_norm": 0.5112618802926924,
+      "learning_rate": 3.366515953829758e-06,
+      "loss": 0.1868,
+      "step": 1855
+    },
+    {
+      "epoch": 0.41817106486045,
+      "grad_norm": 0.5277385827073671,
+      "learning_rate": 3.364768294403455e-06,
+      "loss": 0.1688,
+      "step": 1856
+    },
+    {
+      "epoch": 0.4183963725462585,
+      "grad_norm": 0.4851251657185495,
+      "learning_rate": 3.3630201548035512e-06,
+      "loss": 0.167,
+      "step": 1857
+    },
+    {
+      "epoch": 0.4186216802320669,
+      "grad_norm": 0.4959283834884865,
+      "learning_rate": 3.361271536000723e-06,
+      "loss": 0.1664,
+      "step": 1858
+    },
+    {
+      "epoch": 0.41884698791787534,
+      "grad_norm": 0.4935586642366765,
+      "learning_rate": 3.359522438965915e-06,
+      "loss": 0.164,
+      "step": 1859
+    },
+    {
+      "epoch": 0.41907229560368375,
+      "grad_norm": 0.46178173846204396,
+      "learning_rate": 3.3577728646703335e-06,
+      "loss": 0.1493,
+      "step": 1860
+    },
+    {
+      "epoch": 0.4192976032894922,
+      "grad_norm": 0.5079501676323284,
+      "learning_rate": 3.3560228140854534e-06,
+      "loss": 0.165,
+      "step": 1861
+    },
+    {
+      "epoch": 0.41952291097530064,
+      "grad_norm": 0.4762264209677515,
+      "learning_rate": 3.354272288183012e-06,
+      "loss": 0.1665,
+      "step": 1862
+    },
+    {
+      "epoch": 0.41974821866110906,
+      "grad_norm": 0.4572713716633099,
+      "learning_rate": 3.352521287935011e-06,
+      "loss": 0.1449,
+      "step": 1863
+    },
+    {
+      "epoch": 0.41997352634691754,
+      "grad_norm": 0.4898671317124843,
+      "learning_rate": 3.3507698143137157e-06,
+      "loss": 0.1559,
+      "step": 1864
+    },
+    {
+      "epoch": 0.42019883403272595,
+      "grad_norm": 0.48000831024326635,
+      "learning_rate": 3.3490178682916534e-06,
+      "loss": 0.1443,
+      "step": 1865
+    },
+    {
+      "epoch": 0.42042414171853437,
+      "grad_norm": 0.4811857384241612,
+      "learning_rate": 3.3472654508416157e-06,
+      "loss": 0.1605,
+      "step": 1866
+    },
+    {
+      "epoch": 0.4206494494043428,
+      "grad_norm": 0.4738425883508626,
+      "learning_rate": 3.3455125629366546e-06,
+      "loss": 0.1587,
+      "step": 1867
+    },
+    {
+      "epoch": 0.42087475709015126,
+      "grad_norm": 0.47907204424926936,
+      "learning_rate": 3.3437592055500825e-06,
+      "loss": 0.1643,
+      "step": 1868
+    },
+    {
+      "epoch": 0.4211000647759597,
+      "grad_norm": 0.49431849349421114,
+      "learning_rate": 3.342005379655474e-06,
+      "loss": 0.1578,
+      "step": 1869
+    },
+    {
+      "epoch": 0.4213253724617681,
+      "grad_norm": 0.4829550989570886,
+      "learning_rate": 3.340251086226663e-06,
+      "loss": 0.1594,
+      "step": 1870
+    },
+    {
+      "epoch": 0.4215506801475765,
+      "grad_norm": 0.48055300095473713,
+      "learning_rate": 3.3384963262377434e-06,
+      "loss": 0.165,
+      "step": 1871
+    },
+    {
+      "epoch": 0.421775987833385,
+      "grad_norm": 0.5002044469517478,
+      "learning_rate": 3.3367411006630677e-06,
+      "loss": 0.1699,
+      "step": 1872
+    },
+    {
+      "epoch": 0.4220012955191934,
+      "grad_norm": 0.494112142272217,
+      "learning_rate": 3.3349854104772476e-06,
+      "loss": 0.1703,
+      "step": 1873
+    },
+    {
+      "epoch": 0.4222266032050018,
+      "grad_norm": 0.4407900454940755,
+      "learning_rate": 3.333229256655153e-06,
+      "loss": 0.1502,
+      "step": 1874
+    },
+    {
+      "epoch": 0.42245191089081024,
+      "grad_norm": 0.4658230976699728,
+      "learning_rate": 3.3314726401719088e-06,
+      "loss": 0.1503,
+      "step": 1875
+    },
+    {
+      "epoch": 0.4226772185766187,
+      "grad_norm": 0.46842504862609974,
+      "learning_rate": 3.3297155620029e-06,
+      "loss": 0.1508,
+      "step": 1876
+    },
+    {
+      "epoch": 0.42290252626242714,
+      "grad_norm": 0.49870166945340855,
+      "learning_rate": 3.3279580231237664e-06,
+      "loss": 0.1581,
+      "step": 1877
+    },
+    {
+      "epoch": 0.42312783394823555,
+      "grad_norm": 0.4689208756074871,
+      "learning_rate": 3.326200024510405e-06,
+      "loss": 0.1552,
+      "step": 1878
+    },
+    {
+      "epoch": 0.423353141634044,
+      "grad_norm": 0.47798369467976537,
+      "learning_rate": 3.324441567138965e-06,
+      "loss": 0.1632,
+      "step": 1879
+    },
+    {
+      "epoch": 0.42357844931985245,
+      "grad_norm": 0.4426346717785938,
+      "learning_rate": 3.3226826519858526e-06,
+      "loss": 0.1472,
+      "step": 1880
+    },
+    {
+      "epoch": 0.42380375700566086,
+      "grad_norm": 0.45722379581080347,
+      "learning_rate": 3.320923280027728e-06,
+      "loss": 0.1565,
+      "step": 1881
+    },
+    {
+      "epoch": 0.4240290646914693,
+      "grad_norm": 0.47583545692139984,
+      "learning_rate": 3.3191634522415064e-06,
+      "loss": 0.1466,
+      "step": 1882
+    },
+    {
+      "epoch": 0.4242543723772777,
+      "grad_norm": 0.47819718631976604,
+      "learning_rate": 3.317403169604352e-06,
+      "loss": 0.1597,
+      "step": 1883
+    },
+    {
+      "epoch": 0.4244796800630862,
+      "grad_norm": 0.44056899985102965,
+      "learning_rate": 3.315642433093686e-06,
+      "loss": 0.1424,
+      "step": 1884
+    },
+    {
+      "epoch": 0.4247049877488946,
+      "grad_norm": 0.4663525404770331,
+      "learning_rate": 3.313881243687179e-06,
+      "loss": 0.1579,
+      "step": 1885
+    },
+    {
+      "epoch": 0.424930295434703,
+      "grad_norm": 0.49526535778153447,
+      "learning_rate": 3.3121196023627543e-06,
+      "loss": 0.1705,
+      "step": 1886
+    },
+    {
+      "epoch": 0.4251556031205114,
+      "grad_norm": 0.49666092172588805,
+      "learning_rate": 3.3103575100985852e-06,
+      "loss": 0.1592,
+      "step": 1887
+    },
+    {
+      "epoch": 0.4253809108063199,
+      "grad_norm": 0.45343350200228694,
+      "learning_rate": 3.3085949678730953e-06,
+      "loss": 0.1417,
+      "step": 1888
+    },
+    {
+      "epoch": 0.4256062184921283,
+      "grad_norm": 0.45820729838898594,
+      "learning_rate": 3.3068319766649605e-06,
+      "loss": 0.1433,
+      "step": 1889
+    },
+    {
+      "epoch": 0.42583152617793674,
+      "grad_norm": 0.49477542929726775,
+      "learning_rate": 3.305068537453102e-06,
+      "loss": 0.1659,
+      "step": 1890
+    },
+    {
+      "epoch": 0.42605683386374515,
+      "grad_norm": 0.4644258024347816,
+      "learning_rate": 3.303304651216693e-06,
+      "loss": 0.1537,
+      "step": 1891
+    },
+    {
+      "epoch": 0.42628214154955363,
+      "grad_norm": 0.4965959281033501,
+      "learning_rate": 3.3015403189351536e-06,
+      "loss": 0.1714,
+      "step": 1892
+    },
+    {
+      "epoch": 0.42650744923536205,
+      "grad_norm": 0.467098799532159,
+      "learning_rate": 3.2997755415881516e-06,
+      "loss": 0.1511,
+      "step": 1893
+    },
+    {
+      "epoch": 0.42673275692117046,
+      "grad_norm": 0.47792108039203884,
+      "learning_rate": 3.2980103201556023e-06,
+      "loss": 0.1672,
+      "step": 1894
+    },
+    {
+      "epoch": 0.4269580646069789,
+      "grad_norm": 0.5093402711827806,
+      "learning_rate": 3.2962446556176676e-06,
+      "loss": 0.1731,
+      "step": 1895
+    },
+    {
+      "epoch": 0.42718337229278736,
+      "grad_norm": 0.48714060411006055,
+      "learning_rate": 3.2944785489547544e-06,
+      "loss": 0.1524,
+      "step": 1896
+    },
+    {
+      "epoch": 0.4274086799785958,
+      "grad_norm": 0.45516129520081977,
+      "learning_rate": 3.2927120011475168e-06,
+      "loss": 0.1544,
+      "step": 1897
+    },
+    {
+      "epoch": 0.4276339876644042,
+      "grad_norm": 0.4746908996355466,
+      "learning_rate": 3.290945013176852e-06,
+      "loss": 0.1581,
+      "step": 1898
+    },
+    {
+      "epoch": 0.4278592953502126,
+      "grad_norm": 0.44735001802092467,
+      "learning_rate": 3.2891775860239033e-06,
+      "loss": 0.143,
+      "step": 1899
+    },
+    {
+      "epoch": 0.4280846030360211,
+      "grad_norm": 0.5027516078080247,
+      "learning_rate": 3.2874097206700566e-06,
+      "loss": 0.1722,
+      "step": 1900
+    },
+    {
+      "epoch": 0.4283099107218295,
+      "grad_norm": 0.5060602149207811,
+      "learning_rate": 3.285641418096942e-06,
+      "loss": 0.1714,
+      "step": 1901
+    },
+    {
+      "epoch": 0.4285352184076379,
+      "grad_norm": 0.4653823592371517,
+      "learning_rate": 3.2838726792864315e-06,
+      "loss": 0.1542,
+      "step": 1902
+    },
+    {
+      "epoch": 0.42876052609344634,
+      "grad_norm": 0.46278383982423194,
+      "learning_rate": 3.2821035052206413e-06,
+      "loss": 0.1572,
+      "step": 1903
+    },
+    {
+      "epoch": 0.4289858337792548,
+      "grad_norm": 0.4544680300853722,
+      "learning_rate": 3.2803338968819264e-06,
+      "loss": 0.1516,
+      "step": 1904
+    },
+    {
+      "epoch": 0.42921114146506323,
+      "grad_norm": 0.49859338344706805,
+      "learning_rate": 3.278563855252885e-06,
+      "loss": 0.1641,
+      "step": 1905
+    },
+    {
+      "epoch": 0.42943644915087165,
+      "grad_norm": 0.45697671851825894,
+      "learning_rate": 3.2767933813163542e-06,
+      "loss": 0.1446,
+      "step": 1906
+    },
+    {
+      "epoch": 0.42966175683668006,
+      "grad_norm": 0.4505096587808184,
+      "learning_rate": 3.2750224760554135e-06,
+      "loss": 0.155,
+      "step": 1907
+    },
+    {
+      "epoch": 0.42988706452248854,
+      "grad_norm": 0.48709353813394207,
+      "learning_rate": 3.2732511404533797e-06,
+      "loss": 0.1617,
+      "step": 1908
+    },
+    {
+      "epoch": 0.43011237220829696,
+      "grad_norm": 0.518531427247498,
+      "learning_rate": 3.2714793754938102e-06,
+      "loss": 0.1773,
+      "step": 1909
+    },
+    {
+      "epoch": 0.4303376798941054,
+      "grad_norm": 0.46946154006491075,
+      "learning_rate": 3.2697071821604986e-06,
+      "loss": 0.1625,
+      "step": 1910
+    },
+    {
+      "epoch": 0.43056298757991385,
+      "grad_norm": 0.49032743378244426,
+      "learning_rate": 3.2679345614374802e-06,
+      "loss": 0.1733,
+      "step": 1911
+    },
+    {
+      "epoch": 0.43078829526572227,
+      "grad_norm": 0.4637403507116278,
+      "learning_rate": 3.266161514309023e-06,
+      "loss": 0.161,
+      "step": 1912
+    },
+    {
+      "epoch": 0.4310136029515307,
+      "grad_norm": 0.494990037196409,
+      "learning_rate": 3.264388041759635e-06,
+      "loss": 0.1688,
+      "step": 1913
+    },
+    {
+      "epoch": 0.4312389106373391,
+      "grad_norm": 0.4745189714599409,
+      "learning_rate": 3.262614144774059e-06,
+      "loss": 0.1678,
+      "step": 1914
+    },
+    {
+      "epoch": 0.4314642183231476,
+      "grad_norm": 0.44797870354627173,
+      "learning_rate": 3.260839824337274e-06,
+      "loss": 0.1442,
+      "step": 1915
+    },
+    {
+      "epoch": 0.431689526008956,
+      "grad_norm": 0.4667151491338866,
+      "learning_rate": 3.259065081434495e-06,
+      "loss": 0.1518,
+      "step": 1916
+    },
+    {
+      "epoch": 0.4319148336947644,
+      "grad_norm": 0.5028061583487423,
+      "learning_rate": 3.2572899170511683e-06,
+      "loss": 0.1644,
+      "step": 1917
+    },
+    {
+      "epoch": 0.43214014138057283,
+      "grad_norm": 0.4745421286227182,
+      "learning_rate": 3.255514332172979e-06,
+      "loss": 0.1643,
+      "step": 1918
+    },
+    {
+      "epoch": 0.4323654490663813,
+      "grad_norm": 0.4621433702702943,
+      "learning_rate": 3.2537383277858413e-06,
+      "loss": 0.1483,
+      "step": 1919
+    },
+    {
+      "epoch": 0.4325907567521897,
+      "grad_norm": 0.4689770686079878,
+      "learning_rate": 3.2519619048759056e-06,
+      "loss": 0.153,
+      "step": 1920
+    },
+    {
+      "epoch": 0.43281606443799814,
+      "grad_norm": 0.4866272597635595,
+      "learning_rate": 3.250185064429552e-06,
+      "loss": 0.1671,
+      "step": 1921
+    },
+    {
+      "epoch": 0.43304137212380656,
+      "grad_norm": 0.5005319571358561,
+      "learning_rate": 3.248407807433396e-06,
+      "loss": 0.1702,
+      "step": 1922
+    },
+    {
+      "epoch": 0.43326667980961503,
+      "grad_norm": 0.4922067763016835,
+      "learning_rate": 3.246630134874279e-06,
+      "loss": 0.1593,
+      "step": 1923
+    },
+    {
+      "epoch": 0.43349198749542345,
+      "grad_norm": 0.515609895521259,
+      "learning_rate": 3.2448520477392788e-06,
+      "loss": 0.1739,
+      "step": 1924
+    },
+    {
+      "epoch": 0.43371729518123187,
+      "grad_norm": 0.469959696046339,
+      "learning_rate": 3.2430735470157e-06,
+      "loss": 0.1547,
+      "step": 1925
+    },
+    {
+      "epoch": 0.4339426028670403,
+      "grad_norm": 0.4950915860353573,
+      "learning_rate": 3.2412946336910778e-06,
+      "loss": 0.1729,
+      "step": 1926
+    },
+    {
+      "epoch": 0.43416791055284876,
+      "grad_norm": 0.47469229450105466,
+      "learning_rate": 3.2395153087531767e-06,
+      "loss": 0.1738,
+      "step": 1927
+    },
+    {
+      "epoch": 0.4343932182386572,
+      "grad_norm": 0.4894232593544117,
+      "learning_rate": 3.237735573189989e-06,
+      "loss": 0.1675,
+      "step": 1928
+    },
+    {
+      "epoch": 0.4346185259244656,
+      "grad_norm": 0.4519000417682082,
+      "learning_rate": 3.2359554279897353e-06,
+      "loss": 0.1393,
+      "step": 1929
+    },
+    {
+      "epoch": 0.434843833610274,
+      "grad_norm": 0.4904620821483386,
+      "learning_rate": 3.234174874140866e-06,
+      "loss": 0.1566,
+      "step": 1930
+    },
+    {
+      "epoch": 0.4350691412960825,
+      "grad_norm": 0.47125494694001646,
+      "learning_rate": 3.232393912632054e-06,
+      "loss": 0.1527,
+      "step": 1931
+    },
+    {
+      "epoch": 0.4352944489818909,
+      "grad_norm": 0.47128696022337324,
+      "learning_rate": 3.230612544452202e-06,
+      "loss": 0.1644,
+      "step": 1932
+    },
+    {
+      "epoch": 0.4355197566676993,
+      "grad_norm": 0.46240632956038014,
+      "learning_rate": 3.228830770590436e-06,
+      "loss": 0.1537,
+      "step": 1933
+    },
+    {
+      "epoch": 0.43574506435350774,
+      "grad_norm": 0.4480412958017911,
+      "learning_rate": 3.2270485920361093e-06,
+      "loss": 0.1553,
+      "step": 1934
+    },
+    {
+      "epoch": 0.4359703720393162,
+      "grad_norm": 0.47450609805560406,
+      "learning_rate": 3.2252660097788003e-06,
+      "loss": 0.1692,
+      "step": 1935
+    },
+    {
+      "epoch": 0.43619567972512463,
+      "grad_norm": 0.5016631632234949,
+      "learning_rate": 3.2234830248083095e-06,
+      "loss": 0.1612,
+      "step": 1936
+    },
+    {
+      "epoch": 0.43642098741093305,
+      "grad_norm": 0.4630317566529645,
+      "learning_rate": 3.2216996381146613e-06,
+      "loss": 0.1501,
+      "step": 1937
+    },
+    {
+      "epoch": 0.43664629509674147,
+      "grad_norm": 0.4350896225829126,
+      "learning_rate": 3.219915850688106e-06,
+      "loss": 0.1385,
+      "step": 1938
+    },
+    {
+      "epoch": 0.43687160278254994,
+      "grad_norm": 0.4693673421131627,
+      "learning_rate": 3.2181316635191125e-06,
+      "loss": 0.1546,
+      "step": 1939
+    },
+    {
+      "epoch": 0.43709691046835836,
+      "grad_norm": 0.4718115830527021,
+      "learning_rate": 3.2163470775983733e-06,
+      "loss": 0.1505,
+      "step": 1940
+    },
+    {
+      "epoch": 0.4373222181541668,
+      "grad_norm": 0.5096285305771173,
+      "learning_rate": 3.2145620939168036e-06,
+      "loss": 0.1643,
+      "step": 1941
+    },
+    {
+      "epoch": 0.4375475258399752,
+      "grad_norm": 0.4902992882458936,
+      "learning_rate": 3.2127767134655374e-06,
+      "loss": 0.1654,
+      "step": 1942
+    },
+    {
+      "epoch": 0.43777283352578367,
+      "grad_norm": 0.47117853274605337,
+      "learning_rate": 3.210990937235931e-06,
+      "loss": 0.1487,
+      "step": 1943
+    },
+    {
+      "epoch": 0.4379981412115921,
+      "grad_norm": 0.4991686208300184,
+      "learning_rate": 3.209204766219558e-06,
+      "loss": 0.1675,
+      "step": 1944
+    },
+    {
+      "epoch": 0.4382234488974005,
+      "grad_norm": 0.4538891115941367,
+      "learning_rate": 3.207418201408213e-06,
+      "loss": 0.1471,
+      "step": 1945
+    },
+    {
+      "epoch": 0.4384487565832089,
+      "grad_norm": 0.4709571337411855,
+      "learning_rate": 3.205631243793909e-06,
+      "loss": 0.1538,
+      "step": 1946
+    },
+    {
+      "epoch": 0.4386740642690174,
+      "grad_norm": 0.4860122274389726,
+      "learning_rate": 3.2038438943688777e-06,
+      "loss": 0.1653,
+      "step": 1947
+    },
+    {
+      "epoch": 0.4388993719548258,
+      "grad_norm": 0.5018583708691458,
+      "learning_rate": 3.202056154125567e-06,
+      "loss": 0.1632,
+      "step": 1948
+    },
+    {
+      "epoch": 0.43912467964063423,
+      "grad_norm": 0.44754173977669515,
+      "learning_rate": 3.2002680240566412e-06,
+      "loss": 0.1513,
+      "step": 1949
+    },
+    {
+      "epoch": 0.43934998732644265,
+      "grad_norm": 0.45793980293365255,
+      "learning_rate": 3.198479505154984e-06,
+      "loss": 0.1611,
+      "step": 1950
+    },
+    {
+      "epoch": 0.4395752950122511,
+      "grad_norm": 0.4798406577992235,
+      "learning_rate": 3.1966905984136932e-06,
+      "loss": 0.1615,
+      "step": 1951
+    },
+    {
+      "epoch": 0.43980060269805954,
+      "grad_norm": 0.5049353027884466,
+      "learning_rate": 3.1949013048260813e-06,
+      "loss": 0.1739,
+      "step": 1952
+    },
+    {
+      "epoch": 0.44002591038386796,
+      "grad_norm": 0.4691288105733682,
+      "learning_rate": 3.1931116253856762e-06,
+      "loss": 0.1603,
+      "step": 1953
+    },
+    {
+      "epoch": 0.4402512180696764,
+      "grad_norm": 0.48549399893117434,
+      "learning_rate": 3.1913215610862208e-06,
+      "loss": 0.1643,
+      "step": 1954
+    },
+    {
+      "epoch": 0.44047652575548485,
+      "grad_norm": 0.46413359860802367,
+      "learning_rate": 3.189531112921671e-06,
+      "loss": 0.157,
+      "step": 1955
+    },
+    {
+      "epoch": 0.44070183344129327,
+      "grad_norm": 0.48570841885505445,
+      "learning_rate": 3.1877402818861954e-06,
+      "loss": 0.1357,
+      "step": 1956
+    },
+    {
+      "epoch": 0.4409271411271017,
+      "grad_norm": 0.4779260667177893,
+      "learning_rate": 3.185949068974177e-06,
+      "loss": 0.158,
+      "step": 1957
+    },
+    {
+      "epoch": 0.4411524488129101,
+      "grad_norm": 0.4418919545366911,
+      "learning_rate": 3.184157475180208e-06,
+      "loss": 0.1526,
+      "step": 1958
+    },
+    {
+      "epoch": 0.4413777564987186,
+      "grad_norm": 0.49614129483843544,
+      "learning_rate": 3.1823655014990937e-06,
+      "loss": 0.1608,
+      "step": 1959
+    },
+    {
+      "epoch": 0.441603064184527,
+      "grad_norm": 0.472344591085106,
+      "learning_rate": 3.1805731489258516e-06,
+      "loss": 0.1479,
+      "step": 1960
+    },
+    {
+      "epoch": 0.4418283718703354,
+      "grad_norm": 0.49519200575357003,
+      "learning_rate": 3.1787804184557074e-06,
+      "loss": 0.1498,
+      "step": 1961
+    },
+    {
+      "epoch": 0.4420536795561439,
+      "grad_norm": 0.4767514542436889,
+      "learning_rate": 3.1769873110840977e-06,
+      "loss": 0.1567,
+      "step": 1962
+    },
+    {
+      "epoch": 0.4422789872419523,
+      "grad_norm": 0.4586959627133704,
+      "learning_rate": 3.1751938278066687e-06,
+      "loss": 0.1556,
+      "step": 1963
+    },
+    {
+      "epoch": 0.4425042949277607,
+      "grad_norm": 0.4973373090100224,
+      "learning_rate": 3.1733999696192736e-06,
+      "loss": 0.1614,
+      "step": 1964
+    },
+    {
+      "epoch": 0.44272960261356914,
+      "grad_norm": 0.4735446280299149,
+      "learning_rate": 3.171605737517976e-06,
+      "loss": 0.1647,
+      "step": 1965
+    },
+    {
+      "epoch": 0.4429549102993776,
+      "grad_norm": 0.44265387795324745,
+      "learning_rate": 3.1698111324990454e-06,
+      "loss": 0.1407,
+      "step": 1966
+    },
+    {
+      "epoch": 0.44318021798518603,
+      "grad_norm": 0.49122833299235197,
+      "learning_rate": 3.16801615555896e-06,
+      "loss": 0.1684,
+      "step": 1967
+    },
+    {
+      "epoch": 0.44340552567099445,
+      "grad_norm": 0.464417443256395,
+      "learning_rate": 3.1662208076944027e-06,
+      "loss": 0.1543,
+      "step": 1968
+    },
+    {
+      "epoch": 0.44363083335680287,
+      "grad_norm": 0.4593154910225855,
+      "learning_rate": 3.1644250899022637e-06,
+      "loss": 0.1513,
+      "step": 1969
+    },
+    {
+      "epoch": 0.44385614104261134,
+      "grad_norm": 0.47599458719846477,
+      "learning_rate": 3.162629003179638e-06,
+      "loss": 0.1545,
+      "step": 1970
+    },
+    {
+      "epoch": 0.44408144872841976,
+      "grad_norm": 0.48647770396183504,
+      "learning_rate": 3.1608325485238257e-06,
+      "loss": 0.1526,
+      "step": 1971
+    },
+    {
+      "epoch": 0.4443067564142282,
+      "grad_norm": 0.48806006529062335,
+      "learning_rate": 3.1590357269323312e-06,
+      "loss": 0.1641,
+      "step": 1972
+    },
+    {
+      "epoch": 0.4445320641000366,
+      "grad_norm": 0.4838101728349616,
+      "learning_rate": 3.157238539402862e-06,
+      "loss": 0.1691,
+      "step": 1973
+    },
+    {
+      "epoch": 0.44475737178584507,
+      "grad_norm": 0.47238245788795774,
+      "learning_rate": 3.15544098693333e-06,
+      "loss": 0.1707,
+      "step": 1974
+    },
+    {
+      "epoch": 0.4449826794716535,
+      "grad_norm": 0.45538517824292013,
+      "learning_rate": 3.15364307052185e-06,
+      "loss": 0.151,
+      "step": 1975
+    },
+    {
+      "epoch": 0.4452079871574619,
+      "grad_norm": 0.4696391452617115,
+      "learning_rate": 3.151844791166735e-06,
+      "loss": 0.1637,
+      "step": 1976
+    },
+    {
+      "epoch": 0.4454332948432703,
+      "grad_norm": 0.519697876199234,
+      "learning_rate": 3.1500461498665053e-06,
+      "loss": 0.1779,
+      "step": 1977
+    },
+    {
+      "epoch": 0.4456586025290788,
+      "grad_norm": 0.4721751150534047,
+      "learning_rate": 3.1482471476198784e-06,
+      "loss": 0.1598,
+      "step": 1978
+    },
+    {
+      "epoch": 0.4458839102148872,
+      "grad_norm": 0.46400557490164673,
+      "learning_rate": 3.1464477854257726e-06,
+      "loss": 0.1527,
+      "step": 1979
+    },
+    {
+      "epoch": 0.44610921790069563,
+      "grad_norm": 0.47922627381109123,
+      "learning_rate": 3.1446480642833077e-06,
+      "loss": 0.1593,
+      "step": 1980
+    },
+    {
+      "epoch": 0.44633452558650405,
+      "grad_norm": 0.4892235358343949,
+      "learning_rate": 3.1428479851918014e-06,
+      "loss": 0.1602,
+      "step": 1981
+    },
+    {
+      "epoch": 0.4465598332723125,
+      "grad_norm": 0.49009997683912404,
+      "learning_rate": 3.14104754915077e-06,
+      "loss": 0.1576,
+      "step": 1982
+    },
+    {
+      "epoch": 0.44678514095812094,
+      "grad_norm": 0.48069166309580474,
+      "learning_rate": 3.1392467571599288e-06,
+      "loss": 0.1653,
+      "step": 1983
+    },
+    {
+      "epoch": 0.44701044864392936,
+      "grad_norm": 0.4771441161006696,
+      "learning_rate": 3.137445610219192e-06,
+      "loss": 0.1515,
+      "step": 1984
+    },
+    {
+      "epoch": 0.4472357563297378,
+      "grad_norm": 0.4818283293635933,
+      "learning_rate": 3.1356441093286673e-06,
+      "loss": 0.166,
+      "step": 1985
+    },
+    {
+      "epoch": 0.44746106401554625,
+      "grad_norm": 0.5147787178532238,
+      "learning_rate": 3.133842255488661e-06,
+      "loss": 0.174,
+      "step": 1986
+    },
+    {
+      "epoch": 0.44768637170135467,
+      "grad_norm": 0.5344233693465625,
+      "learning_rate": 3.132040049699676e-06,
+      "loss": 0.1763,
+      "step": 1987
+    },
+    {
+      "epoch": 0.4479116793871631,
+      "grad_norm": 0.48389319685884197,
+      "learning_rate": 3.130237492962411e-06,
+      "loss": 0.1643,
+      "step": 1988
+    },
+    {
+      "epoch": 0.4481369870729715,
+      "grad_norm": 0.482536320658805,
+      "learning_rate": 3.1284345862777572e-06,
+      "loss": 0.1595,
+      "step": 1989
+    },
+    {
+      "epoch": 0.44836229475878,
+      "grad_norm": 0.46393789436271843,
+      "learning_rate": 3.1266313306468018e-06,
+      "loss": 0.1524,
+      "step": 1990
+    },
+    {
+      "epoch": 0.4485876024445884,
+      "grad_norm": 0.474674964893104,
+      "learning_rate": 3.1248277270708255e-06,
+      "loss": 0.1601,
+      "step": 1991
+    },
+    {
+      "epoch": 0.4488129101303968,
+      "grad_norm": 0.490753280567544,
+      "learning_rate": 3.1230237765513023e-06,
+      "loss": 0.1721,
+      "step": 1992
+    },
+    {
+      "epoch": 0.44903821781620523,
+      "grad_norm": 0.4499975991773963,
+      "learning_rate": 3.121219480089899e-06,
+      "loss": 0.1446,
+      "step": 1993
+    },
+    {
+      "epoch": 0.4492635255020137,
+      "grad_norm": 0.458297504071037,
+      "learning_rate": 3.119414838688473e-06,
+      "loss": 0.1529,
+      "step": 1994
+    },
+    {
+      "epoch": 0.4494888331878221,
+      "grad_norm": 0.4436947744388515,
+      "learning_rate": 3.1176098533490755e-06,
+      "loss": 0.1386,
+      "step": 1995
+    },
+    {
+      "epoch": 0.44971414087363054,
+      "grad_norm": 0.5104688488260797,
+      "learning_rate": 3.1158045250739473e-06,
+      "loss": 0.1746,
+      "step": 1996
+    },
+    {
+      "epoch": 0.44993944855943896,
+      "grad_norm": 0.4881255207930648,
+      "learning_rate": 3.11399885486552e-06,
+      "loss": 0.1533,
+      "step": 1997
+    },
+    {
+      "epoch": 0.45016475624524743,
+      "grad_norm": 0.49185335203111147,
+      "learning_rate": 3.1121928437264138e-06,
+      "loss": 0.1572,
+      "step": 1998
+    },
+    {
+      "epoch": 0.45039006393105585,
+      "grad_norm": 0.4923412831035321,
+      "learning_rate": 3.1103864926594406e-06,
+      "loss": 0.1532,
+      "step": 1999
+    },
+    {
+      "epoch": 0.45061537161686427,
+      "grad_norm": 0.48052935278952436,
+      "learning_rate": 3.1085798026676e-06,
+      "loss": 0.1557,
+      "step": 2000
+    },
+    {
+      "epoch": 0.45061537161686427,
+      "eval_loss": 0.15805742144584656,
+      "eval_runtime": 56.952,
+      "eval_samples_per_second": 50.393,
+      "eval_steps_per_second": 6.304,
+      "step": 2000
+    },
+    {
+      "epoch": 0.4508406793026727,
+      "grad_norm": 0.46526999043209805,
+      "learning_rate": 3.1067727747540797e-06,
+      "loss": 0.1411,
+      "step": 2001
+    },
+    {
+      "epoch": 0.45106598698848116,
+      "grad_norm": 0.4607371866430295,
+      "learning_rate": 3.1049654099222542e-06,
+      "loss": 0.1581,
+      "step": 2002
+    },
+    {
+      "epoch": 0.4512912946742896,
+      "grad_norm": 0.49592540797717755,
+      "learning_rate": 3.1031577091756852e-06,
+      "loss": 0.1536,
+      "step": 2003
+    },
+    {
+      "epoch": 0.451516602360098,
+      "grad_norm": 0.46818841838662356,
+      "learning_rate": 3.1013496735181232e-06,
+      "loss": 0.1492,
+      "step": 2004
+    },
+    {
+      "epoch": 0.4517419100459064,
+      "grad_norm": 0.46040611469553566,
+      "learning_rate": 3.0995413039535017e-06,
+      "loss": 0.1534,
+      "step": 2005
+    },
+    {
+      "epoch": 0.4519672177317149,
+      "grad_norm": 0.4973561683635516,
+      "learning_rate": 3.0977326014859415e-06,
+      "loss": 0.1725,
+      "step": 2006
+    },
+    {
+      "epoch": 0.4521925254175233,
+      "grad_norm": 0.460212661652161,
+      "learning_rate": 3.095923567119748e-06,
+      "loss": 0.1557,
+      "step": 2007
+    },
+    {
+      "epoch": 0.4524178331033317,
+      "grad_norm": 0.4715234518160938,
+      "learning_rate": 3.09411420185941e-06,
+      "loss": 0.1614,
+      "step": 2008
+    },
+    {
+      "epoch": 0.4526431407891402,
+      "grad_norm": 0.47436206347918625,
+      "learning_rate": 3.0923045067096e-06,
+      "loss": 0.155,
+      "step": 2009
+    },
+    {
+      "epoch": 0.4528684484749486,
+      "grad_norm": 0.48006343097393855,
+      "learning_rate": 3.090494482675176e-06,
+      "loss": 0.1714,
+      "step": 2010
+    },
+    {
+      "epoch": 0.45309375616075703,
+      "grad_norm": 0.48029448201908953,
+      "learning_rate": 3.088684130761175e-06,
+      "loss": 0.1661,
+      "step": 2011
+    },
+    {
+      "epoch": 0.45331906384656545,
+      "grad_norm": 0.4503633138858404,
+      "learning_rate": 3.0868734519728194e-06,
+      "loss": 0.1425,
+      "step": 2012
+    },
+    {
+      "epoch": 0.4535443715323739,
+      "grad_norm": 0.5150747727818801,
+      "learning_rate": 3.085062447315511e-06,
+      "loss": 0.1675,
+      "step": 2013
+    },
+    {
+      "epoch": 0.45376967921818234,
+      "grad_norm": 0.44458727455954306,
+      "learning_rate": 3.0832511177948326e-06,
+      "loss": 0.1468,
+      "step": 2014
+    },
+    {
+      "epoch": 0.45399498690399076,
+      "grad_norm": 0.5048541437741968,
+      "learning_rate": 3.081439464416549e-06,
+      "loss": 0.173,
+      "step": 2015
+    },
+    {
+      "epoch": 0.4542202945897992,
+      "grad_norm": 0.4810433869141697,
+      "learning_rate": 3.0796274881866034e-06,
+      "loss": 0.1647,
+      "step": 2016
+    },
+    {
+      "epoch": 0.45444560227560765,
+      "grad_norm": 0.4669019629530169,
+      "learning_rate": 3.0778151901111187e-06,
+      "loss": 0.1625,
+      "step": 2017
+    },
+    {
+      "epoch": 0.45467090996141607,
+      "grad_norm": 0.44513180434068006,
+      "learning_rate": 3.0760025711963964e-06,
+      "loss": 0.1404,
+      "step": 2018
+    },
+    {
+      "epoch": 0.4548962176472245,
+      "grad_norm": 0.5034063685601008,
+      "learning_rate": 3.0741896324489163e-06,
+      "loss": 0.1702,
+      "step": 2019
+    },
+    {
+      "epoch": 0.4551215253330329,
+      "grad_norm": 0.4557191252849021,
+      "learning_rate": 3.0723763748753354e-06,
+      "loss": 0.1516,
+      "step": 2020
+    },
+    {
+      "epoch": 0.4553468330188414,
+      "grad_norm": 0.48336102411704346,
+      "learning_rate": 3.0705627994824887e-06,
+      "loss": 0.1595,
+      "step": 2021
+    },
+    {
+      "epoch": 0.4555721407046498,
+      "grad_norm": 0.4485419976949585,
+      "learning_rate": 3.0687489072773864e-06,
+      "loss": 0.1499,
+      "step": 2022
+    },
+    {
+      "epoch": 0.4557974483904582,
+      "grad_norm": 0.4615635228199084,
+      "learning_rate": 3.0669346992672156e-06,
+      "loss": 0.1488,
+      "step": 2023
+    },
+    {
+      "epoch": 0.45602275607626663,
+      "grad_norm": 0.47702823681473344,
+      "learning_rate": 3.0651201764593375e-06,
+      "loss": 0.1628,
+      "step": 2024
+    },
+    {
+      "epoch": 0.4562480637620751,
+      "grad_norm": 0.46354927078581676,
+      "learning_rate": 3.06330533986129e-06,
+      "loss": 0.1554,
+      "step": 2025
+    },
+    {
+      "epoch": 0.4564733714478835,
+      "grad_norm": 0.4654409790097615,
+      "learning_rate": 3.0614901904807836e-06,
+      "loss": 0.1536,
+      "step": 2026
+    },
+    {
+      "epoch": 0.45669867913369194,
+      "grad_norm": 0.4515681205108471,
+      "learning_rate": 3.0596747293257047e-06,
+      "loss": 0.154,
+      "step": 2027
+    },
+    {
+      "epoch": 0.45692398681950036,
+      "grad_norm": 0.4789039895573779,
+      "learning_rate": 3.0578589574041097e-06,
+      "loss": 0.1495,
+      "step": 2028
+    },
+    {
+      "epoch": 0.45714929450530883,
+      "grad_norm": 0.470697922964349,
+      "learning_rate": 3.056042875724228e-06,
+      "loss": 0.1676,
+      "step": 2029
+    },
+    {
+      "epoch": 0.45737460219111725,
+      "grad_norm": 0.4563412984778146,
+      "learning_rate": 3.0542264852944635e-06,
+      "loss": 0.145,
+      "step": 2030
+    },
+    {
+      "epoch": 0.45759990987692567,
+      "grad_norm": 0.48369559208181934,
+      "learning_rate": 3.052409787123391e-06,
+      "loss": 0.1634,
+      "step": 2031
+    },
+    {
+      "epoch": 0.4578252175627341,
+      "grad_norm": 0.4564814760737555,
+      "learning_rate": 3.0505927822197533e-06,
+      "loss": 0.1509,
+      "step": 2032
+    },
+    {
+      "epoch": 0.45805052524854256,
+      "grad_norm": 0.4610972303240204,
+      "learning_rate": 3.0487754715924674e-06,
+      "loss": 0.1413,
+      "step": 2033
+    },
+    {
+      "epoch": 0.458275832934351,
+      "grad_norm": 0.5102253098565472,
+      "learning_rate": 3.0469578562506165e-06,
+      "loss": 0.1575,
+      "step": 2034
+    },
+    {
+      "epoch": 0.4585011406201594,
+      "grad_norm": 0.47644473351215494,
+      "learning_rate": 3.045139937203455e-06,
+      "loss": 0.155,
+      "step": 2035
+    },
+    {
+      "epoch": 0.4587264483059678,
+      "grad_norm": 0.4840780941352861,
+      "learning_rate": 3.0433217154604067e-06,
+      "loss": 0.1725,
+      "step": 2036
+    },
+    {
+      "epoch": 0.4589517559917763,
+      "grad_norm": 0.4675495037083089,
+      "learning_rate": 3.0415031920310613e-06,
+      "loss": 0.1578,
+      "step": 2037
+    },
+    {
+      "epoch": 0.4591770636775847,
+      "grad_norm": 0.466053222288305,
+      "learning_rate": 3.0396843679251777e-06,
+      "loss": 0.1557,
+      "step": 2038
+    },
+    {
+      "epoch": 0.4594023713633931,
+      "grad_norm": 0.4658213885418776,
+      "learning_rate": 3.03786524415268e-06,
+      "loss": 0.1509,
+      "step": 2039
+    },
+    {
+      "epoch": 0.45962767904920154,
+      "grad_norm": 0.47486803158127977,
+      "learning_rate": 3.0360458217236604e-06,
+      "loss": 0.158,
+      "step": 2040
+    },
+    {
+      "epoch": 0.45985298673501,
+      "grad_norm": 0.46493147154073844,
+      "learning_rate": 3.034226101648377e-06,
+      "loss": 0.1562,
+      "step": 2041
+    },
+    {
+      "epoch": 0.46007829442081843,
+      "grad_norm": 0.46366054346391106,
+      "learning_rate": 3.0324060849372526e-06,
+      "loss": 0.1481,
+      "step": 2042
+    },
+    {
+      "epoch": 0.46030360210662685,
+      "grad_norm": 0.4471818204754162,
+      "learning_rate": 3.0305857726008736e-06,
+      "loss": 0.1502,
+      "step": 2043
+    },
+    {
+      "epoch": 0.46052890979243527,
+      "grad_norm": 0.45710029387132534,
+      "learning_rate": 3.028765165649992e-06,
+      "loss": 0.1526,
+      "step": 2044
+    },
+    {
+      "epoch": 0.46075421747824374,
+      "grad_norm": 0.4532254215793454,
+      "learning_rate": 3.026944265095524e-06,
+      "loss": 0.1428,
+      "step": 2045
+    },
+    {
+      "epoch": 0.46097952516405216,
+      "grad_norm": 0.4781547388158946,
+      "learning_rate": 3.0251230719485465e-06,
+      "loss": 0.1522,
+      "step": 2046
+    },
+    {
+      "epoch": 0.4612048328498606,
+      "grad_norm": 0.4679788316950286,
+      "learning_rate": 3.0233015872203004e-06,
+      "loss": 0.1549,
+      "step": 2047
+    },
+    {
+      "epoch": 0.461430140535669,
+      "grad_norm": 0.457693832340645,
+      "learning_rate": 3.0214798119221884e-06,
+      "loss": 0.1559,
+      "step": 2048
+    },
+    {
+      "epoch": 0.46165544822147747,
+      "grad_norm": 0.4618043640061542,
+      "learning_rate": 3.0196577470657744e-06,
+      "loss": 0.1507,
+      "step": 2049
+    },
+    {
+      "epoch": 0.4618807559072859,
+      "grad_norm": 0.46599511428045387,
+      "learning_rate": 3.0178353936627835e-06,
+      "loss": 0.1571,
+      "step": 2050
+    },
+    {
+      "epoch": 0.4621060635930943,
+      "grad_norm": 0.48849016131669354,
+      "learning_rate": 3.0160127527250993e-06,
+      "loss": 0.1559,
+      "step": 2051
+    },
+    {
+      "epoch": 0.4623313712789027,
+      "grad_norm": 0.4734928184582273,
+      "learning_rate": 3.0141898252647682e-06,
+      "loss": 0.1518,
+      "step": 2052
+    },
+    {
+      "epoch": 0.4625566789647112,
+      "grad_norm": 0.47462773177763673,
+      "learning_rate": 3.012366612293993e-06,
+      "loss": 0.1595,
+      "step": 2053
+    },
+    {
+      "epoch": 0.4627819866505196,
+      "grad_norm": 0.49632927526910214,
+      "learning_rate": 3.0105431148251364e-06,
+      "loss": 0.168,
+      "step": 2054
+    },
+    {
+      "epoch": 0.46300729433632803,
+      "grad_norm": 0.44508473522061126,
+      "learning_rate": 3.0087193338707175e-06,
+      "loss": 0.1431,
+      "step": 2055
+    },
+    {
+      "epoch": 0.46323260202213645,
+      "grad_norm": 0.4850902705606218,
+      "learning_rate": 3.0068952704434145e-06,
+      "loss": 0.1682,
+      "step": 2056
+    },
+    {
+      "epoch": 0.4634579097079449,
+      "grad_norm": 0.4585047739418811,
+      "learning_rate": 3.0050709255560616e-06,
+      "loss": 0.1492,
+      "step": 2057
+    },
+    {
+      "epoch": 0.46368321739375334,
+      "grad_norm": 0.4730071664371849,
+      "learning_rate": 3.0032463002216504e-06,
+      "loss": 0.1599,
+      "step": 2058
+    },
+    {
+      "epoch": 0.46390852507956176,
+      "grad_norm": 0.47482057227207636,
+      "learning_rate": 3.0014213954533265e-06,
+      "loss": 0.1533,
+      "step": 2059
+    },
+    {
+      "epoch": 0.46413383276537024,
+      "grad_norm": 0.4421130204730638,
+      "learning_rate": 2.999596212264392e-06,
+      "loss": 0.1522,
+      "step": 2060
+    },
+    {
+      "epoch": 0.46435914045117865,
+      "grad_norm": 0.46017986322521726,
+      "learning_rate": 2.997770751668302e-06,
+      "loss": 0.1489,
+      "step": 2061
+    },
+    {
+      "epoch": 0.46458444813698707,
+      "grad_norm": 0.48128701337916485,
+      "learning_rate": 2.9959450146786674e-06,
+      "loss": 0.1647,
+      "step": 2062
+    },
+    {
+      "epoch": 0.4648097558227955,
+      "grad_norm": 0.4892451718733948,
+      "learning_rate": 2.994119002309253e-06,
+      "loss": 0.1684,
+      "step": 2063
+    },
+    {
+      "epoch": 0.46503506350860396,
+      "grad_norm": 0.4825674322636185,
+      "learning_rate": 2.9922927155739737e-06,
+      "loss": 0.1642,
+      "step": 2064
+    },
+    {
+      "epoch": 0.4652603711944124,
+      "grad_norm": 0.4603373858917702,
+      "learning_rate": 2.9904661554868997e-06,
+      "loss": 0.1404,
+      "step": 2065
+    },
+    {
+      "epoch": 0.4654856788802208,
+      "grad_norm": 0.46565245742689776,
+      "learning_rate": 2.9886393230622507e-06,
+      "loss": 0.1427,
+      "step": 2066
+    },
+    {
+      "epoch": 0.4657109865660292,
+      "grad_norm": 0.484180951774229,
+      "learning_rate": 2.986812219314399e-06,
+      "loss": 0.16,
+      "step": 2067
+    },
+    {
+      "epoch": 0.4659362942518377,
+      "grad_norm": 0.4941176192277806,
+      "learning_rate": 2.984984845257868e-06,
+      "loss": 0.1671,
+      "step": 2068
+    },
+    {
+      "epoch": 0.4661616019376461,
+      "grad_norm": 0.4853395880449739,
+      "learning_rate": 2.983157201907329e-06,
+      "loss": 0.1632,
+      "step": 2069
+    },
+    {
+      "epoch": 0.4663869096234545,
+      "grad_norm": 0.484149010323965,
+      "learning_rate": 2.981329290277605e-06,
+      "loss": 0.1486,
+      "step": 2070
+    },
+    {
+      "epoch": 0.46661221730926294,
+      "grad_norm": 0.47891308586143655,
+      "learning_rate": 2.9795011113836686e-06,
+      "loss": 0.1537,
+      "step": 2071
+    },
+    {
+      "epoch": 0.4668375249950714,
+      "grad_norm": 0.4927883402568322,
+      "learning_rate": 2.977672666240636e-06,
+      "loss": 0.1639,
+      "step": 2072
+    },
+    {
+      "epoch": 0.46706283268087984,
+      "grad_norm": 0.4744327591106369,
+      "learning_rate": 2.9758439558637774e-06,
+      "loss": 0.1562,
+      "step": 2073
+    },
+    {
+      "epoch": 0.46728814036668825,
+      "grad_norm": 0.4941847293633856,
+      "learning_rate": 2.974014981268507e-06,
+      "loss": 0.159,
+      "step": 2074
+    },
+    {
+      "epoch": 0.46751344805249667,
+      "grad_norm": 0.4943272692654053,
+      "learning_rate": 2.972185743470386e-06,
+      "loss": 0.1601,
+      "step": 2075
+    },
+    {
+      "epoch": 0.46773875573830515,
+      "grad_norm": 0.46189492924477654,
+      "learning_rate": 2.9703562434851218e-06,
+      "loss": 0.15,
+      "step": 2076
+    },
+    {
+      "epoch": 0.46796406342411356,
+      "grad_norm": 0.4392177203658699,
+      "learning_rate": 2.9685264823285676e-06,
+      "loss": 0.1428,
+      "step": 2077
+    },
+    {
+      "epoch": 0.468189371109922,
+      "grad_norm": 0.4660710812280689,
+      "learning_rate": 2.966696461016721e-06,
+      "loss": 0.1535,
+      "step": 2078
+    },
+    {
+      "epoch": 0.4684146787957304,
+      "grad_norm": 0.4464453798972159,
+      "learning_rate": 2.964866180565725e-06,
+      "loss": 0.1412,
+      "step": 2079
+    },
+    {
+      "epoch": 0.4686399864815389,
+      "grad_norm": 0.47979725171124965,
+      "learning_rate": 2.9630356419918682e-06,
+      "loss": 0.1612,
+      "step": 2080
+    },
+    {
+      "epoch": 0.4688652941673473,
+      "grad_norm": 0.4805922736007531,
+      "learning_rate": 2.9612048463115774e-06,
+      "loss": 0.1488,
+      "step": 2081
+    },
+    {
+      "epoch": 0.4690906018531557,
+      "grad_norm": 0.4956729872123908,
+      "learning_rate": 2.9593737945414264e-06,
+      "loss": 0.1655,
+      "step": 2082
+    },
+    {
+      "epoch": 0.4693159095389641,
+      "grad_norm": 0.4920903518112405,
+      "learning_rate": 2.9575424876981298e-06,
+      "loss": 0.1586,
+      "step": 2083
+    },
+    {
+      "epoch": 0.4695412172247726,
+      "grad_norm": 0.46898887966562713,
+      "learning_rate": 2.9557109267985445e-06,
+      "loss": 0.1447,
+      "step": 2084
+    },
+    {
+      "epoch": 0.469766524910581,
+      "grad_norm": 0.46688066208914697,
+      "learning_rate": 2.953879112859668e-06,
+      "loss": 0.1487,
+      "step": 2085
+    },
+    {
+      "epoch": 0.46999183259638944,
+      "grad_norm": 0.47890920214338073,
+      "learning_rate": 2.952047046898637e-06,
+      "loss": 0.1704,
+      "step": 2086
+    },
+    {
+      "epoch": 0.47021714028219785,
+      "grad_norm": 0.4993549169939433,
+      "learning_rate": 2.9502147299327316e-06,
+      "loss": 0.171,
+      "step": 2087
+    },
+    {
+      "epoch": 0.4704424479680063,
+      "grad_norm": 0.4829359338136954,
+      "learning_rate": 2.9483821629793673e-06,
+      "loss": 0.1616,
+      "step": 2088
+    },
+    {
+      "epoch": 0.47066775565381475,
+      "grad_norm": 0.4999751762418861,
+      "learning_rate": 2.946549347056101e-06,
+      "loss": 0.1657,
+      "step": 2089
+    },
+    {
+      "epoch": 0.47089306333962316,
+      "grad_norm": 0.47745495500884666,
+      "learning_rate": 2.9447162831806275e-06,
+      "loss": 0.1555,
+      "step": 2090
+    },
+    {
+      "epoch": 0.4711183710254316,
+      "grad_norm": 0.46324150350719656,
+      "learning_rate": 2.942882972370778e-06,
+      "loss": 0.1382,
+      "step": 2091
+    },
+    {
+      "epoch": 0.47134367871124006,
+      "grad_norm": 0.47875126249754046,
+      "learning_rate": 2.941049415644522e-06,
+      "loss": 0.1567,
+      "step": 2092
+    },
+    {
+      "epoch": 0.4715689863970485,
+      "grad_norm": 0.48557336281168423,
+      "learning_rate": 2.9392156140199644e-06,
+      "loss": 0.1587,
+      "step": 2093
+    },
+    {
+      "epoch": 0.4717942940828569,
+      "grad_norm": 0.49638521974648225,
+      "learning_rate": 2.9373815685153485e-06,
+      "loss": 0.162,
+      "step": 2094
+    },
+    {
+      "epoch": 0.4720196017686653,
+      "grad_norm": 0.49484643934095707,
+      "learning_rate": 2.93554728014905e-06,
+      "loss": 0.16,
+      "step": 2095
+    },
+    {
+      "epoch": 0.4722449094544738,
+      "grad_norm": 0.46146720058527874,
+      "learning_rate": 2.933712749939582e-06,
+      "loss": 0.1534,
+      "step": 2096
+    },
+    {
+      "epoch": 0.4724702171402822,
+      "grad_norm": 0.4545334918204591,
+      "learning_rate": 2.9318779789055894e-06,
+      "loss": 0.1432,
+      "step": 2097
+    },
+    {
+      "epoch": 0.4726955248260906,
+      "grad_norm": 0.4554966952835028,
+      "learning_rate": 2.9300429680658538e-06,
+      "loss": 0.1435,
+      "step": 2098
+    },
+    {
+      "epoch": 0.47292083251189904,
+      "grad_norm": 0.4724184782115837,
+      "learning_rate": 2.928207718439287e-06,
+      "loss": 0.1518,
+      "step": 2099
+    },
+    {
+      "epoch": 0.4731461401977075,
+      "grad_norm": 0.4845200135364393,
+      "learning_rate": 2.9263722310449353e-06,
+      "loss": 0.1554,
+      "step": 2100
+    },
+    {
+      "epoch": 0.47337144788351593,
+      "grad_norm": 0.46797662996106126,
+      "learning_rate": 2.924536506901976e-06,
+      "loss": 0.1536,
+      "step": 2101
+    },
+    {
+      "epoch": 0.47359675556932435,
+      "grad_norm": 0.47906899237945594,
+      "learning_rate": 2.9227005470297194e-06,
+      "loss": 0.1649,
+      "step": 2102
+    },
+    {
+      "epoch": 0.47382206325513276,
+      "grad_norm": 0.4618144263119681,
+      "learning_rate": 2.9208643524476037e-06,
+      "loss": 0.1518,
+      "step": 2103
+    },
+    {
+      "epoch": 0.47404737094094124,
+      "grad_norm": 0.46748084889828573,
+      "learning_rate": 2.919027924175201e-06,
+      "loss": 0.1492,
+      "step": 2104
+    },
+    {
+      "epoch": 0.47427267862674966,
+      "grad_norm": 0.5056719949637513,
+      "learning_rate": 2.9171912632322102e-06,
+      "loss": 0.1619,
+      "step": 2105
+    },
+    {
+      "epoch": 0.4744979863125581,
+      "grad_norm": 0.48782787073562445,
+      "learning_rate": 2.915354370638462e-06,
+      "loss": 0.1598,
+      "step": 2106
+    },
+    {
+      "epoch": 0.47472329399836655,
+      "grad_norm": 0.47920209045806517,
+      "learning_rate": 2.913517247413914e-06,
+      "loss": 0.1497,
+      "step": 2107
+    },
+    {
+      "epoch": 0.47494860168417496,
+      "grad_norm": 0.469980252183471,
+      "learning_rate": 2.9116798945786515e-06,
+      "loss": 0.1575,
+      "step": 2108
+    },
+    {
+      "epoch": 0.4751739093699834,
+      "grad_norm": 0.4846588971324414,
+      "learning_rate": 2.909842313152888e-06,
+      "loss": 0.1511,
+      "step": 2109
+    },
+    {
+      "epoch": 0.4753992170557918,
+      "grad_norm": 0.5070918931965868,
+      "learning_rate": 2.9080045041569647e-06,
+      "loss": 0.1716,
+      "step": 2110
+    },
+    {
+      "epoch": 0.4756245247416003,
+      "grad_norm": 0.4917914827122371,
+      "learning_rate": 2.9061664686113487e-06,
+      "loss": 0.1559,
+      "step": 2111
+    },
+    {
+      "epoch": 0.4758498324274087,
+      "grad_norm": 0.489691610534915,
+      "learning_rate": 2.904328207536632e-06,
+      "loss": 0.1543,
+      "step": 2112
+    },
+    {
+      "epoch": 0.4760751401132171,
+      "grad_norm": 0.48507920331153476,
+      "learning_rate": 2.9024897219535326e-06,
+      "loss": 0.1653,
+      "step": 2113
+    },
+    {
+      "epoch": 0.47630044779902553,
+      "grad_norm": 0.47914495383302097,
+      "learning_rate": 2.900651012882893e-06,
+      "loss": 0.152,
+      "step": 2114
+    },
+    {
+      "epoch": 0.476525755484834,
+      "grad_norm": 0.49396402566447883,
+      "learning_rate": 2.8988120813456794e-06,
+      "loss": 0.1613,
+      "step": 2115
+    },
+    {
+      "epoch": 0.4767510631706424,
+      "grad_norm": 0.4811425757009818,
+      "learning_rate": 2.896972928362983e-06,
+      "loss": 0.1578,
+      "step": 2116
+    },
+    {
+      "epoch": 0.47697637085645084,
+      "grad_norm": 0.47253258227579886,
+      "learning_rate": 2.8951335549560156e-06,
+      "loss": 0.1679,
+      "step": 2117
+    },
+    {
+      "epoch": 0.47720167854225926,
+      "grad_norm": 0.4684167266249312,
+      "learning_rate": 2.893293962146114e-06,
+      "loss": 0.1509,
+      "step": 2118
+    },
+    {
+      "epoch": 0.47742698622806773,
+      "grad_norm": 0.4615876351385914,
+      "learning_rate": 2.8914541509547345e-06,
+      "loss": 0.1528,
+      "step": 2119
+    },
+    {
+      "epoch": 0.47765229391387615,
+      "grad_norm": 0.4725901325614137,
+      "learning_rate": 2.8896141224034554e-06,
+      "loss": 0.144,
+      "step": 2120
+    },
+    {
+      "epoch": 0.47787760159968456,
+      "grad_norm": 0.47688211275536097,
+      "learning_rate": 2.8877738775139767e-06,
+      "loss": 0.1503,
+      "step": 2121
+    },
+    {
+      "epoch": 0.478102909285493,
+      "grad_norm": 0.461565550780441,
+      "learning_rate": 2.885933417308118e-06,
+      "loss": 0.1392,
+      "step": 2122
+    },
+    {
+      "epoch": 0.47832821697130146,
+      "grad_norm": 0.47986650743199105,
+      "learning_rate": 2.8840927428078185e-06,
+      "loss": 0.151,
+      "step": 2123
+    },
+    {
+      "epoch": 0.4785535246571099,
+      "grad_norm": 0.5102470451322486,
+      "learning_rate": 2.8822518550351356e-06,
+      "loss": 0.177,
+      "step": 2124
+    },
+    {
+      "epoch": 0.4787788323429183,
+      "grad_norm": 0.4810562269804644,
+      "learning_rate": 2.8804107550122453e-06,
+      "loss": 0.1599,
+      "step": 2125
+    },
+    {
+      "epoch": 0.4790041400287267,
+      "grad_norm": 0.49106145138432467,
+      "learning_rate": 2.878569443761442e-06,
+      "loss": 0.1552,
+      "step": 2126
+    },
+    {
+      "epoch": 0.4792294477145352,
+      "grad_norm": 0.4887172792733754,
+      "learning_rate": 2.8767279223051375e-06,
+      "loss": 0.1536,
+      "step": 2127
+    },
+    {
+      "epoch": 0.4794547554003436,
+      "grad_norm": 0.4689500997214312,
+      "learning_rate": 2.87488619166586e-06,
+      "loss": 0.1549,
+      "step": 2128
+    },
+    {
+      "epoch": 0.479680063086152,
+      "grad_norm": 0.45929306443811646,
+      "learning_rate": 2.8730442528662537e-06,
+      "loss": 0.1614,
+      "step": 2129
+    },
+    {
+      "epoch": 0.47990537077196044,
+      "grad_norm": 0.4871163529397986,
+      "learning_rate": 2.8712021069290786e-06,
+      "loss": 0.1544,
+      "step": 2130
+    },
+    {
+      "epoch": 0.4801306784577689,
+      "grad_norm": 0.4817155544257701,
+      "learning_rate": 2.869359754877209e-06,
+      "loss": 0.1588,
+      "step": 2131
+    },
+    {
+      "epoch": 0.48035598614357733,
+      "grad_norm": 0.4484801331749894,
+      "learning_rate": 2.8675171977336357e-06,
+      "loss": 0.1442,
+      "step": 2132
+    },
+    {
+      "epoch": 0.48058129382938575,
+      "grad_norm": 0.4641674894247696,
+      "learning_rate": 2.8656744365214622e-06,
+      "loss": 0.1528,
+      "step": 2133
+    },
+    {
+      "epoch": 0.48080660151519417,
+      "grad_norm": 0.5012156150664434,
+      "learning_rate": 2.863831472263904e-06,
+      "loss": 0.1566,
+      "step": 2134
+    },
+    {
+      "epoch": 0.48103190920100264,
+      "grad_norm": 0.4633341637924984,
+      "learning_rate": 2.8619883059842897e-06,
+      "loss": 0.1427,
+      "step": 2135
+    },
+    {
+      "epoch": 0.48125721688681106,
+      "grad_norm": 0.4797143088628369,
+      "learning_rate": 2.8601449387060622e-06,
+      "loss": 0.161,
+      "step": 2136
+    },
+    {
+      "epoch": 0.4814825245726195,
+      "grad_norm": 0.47702027421937043,
+      "learning_rate": 2.858301371452774e-06,
+      "loss": 0.1428,
+      "step": 2137
+    },
+    {
+      "epoch": 0.4817078322584279,
+      "grad_norm": 0.5015112965273413,
+      "learning_rate": 2.8564576052480895e-06,
+      "loss": 0.168,
+      "step": 2138
+    },
+    {
+      "epoch": 0.48193313994423637,
+      "grad_norm": 0.4822255713488226,
+      "learning_rate": 2.8546136411157843e-06,
+      "loss": 0.1541,
+      "step": 2139
+    },
+    {
+      "epoch": 0.4821584476300448,
+      "grad_norm": 0.465307141222416,
+      "learning_rate": 2.8527694800797417e-06,
+      "loss": 0.1619,
+      "step": 2140
+    },
+    {
+      "epoch": 0.4823837553158532,
+      "grad_norm": 0.50266265761066,
+      "learning_rate": 2.850925123163956e-06,
+      "loss": 0.1698,
+      "step": 2141
+    },
+    {
+      "epoch": 0.4826090630016616,
+      "grad_norm": 0.4421128901521324,
+      "learning_rate": 2.8490805713925298e-06,
+      "loss": 0.1435,
+      "step": 2142
+    },
+    {
+      "epoch": 0.4828343706874701,
+      "grad_norm": 0.4812431803950061,
+      "learning_rate": 2.847235825789673e-06,
+      "loss": 0.1539,
+      "step": 2143
+    },
+    {
+      "epoch": 0.4830596783732785,
+      "grad_norm": 0.4818346965554266,
+      "learning_rate": 2.845390887379706e-06,
+      "loss": 0.1687,
+      "step": 2144
+    },
+    {
+      "epoch": 0.48328498605908693,
+      "grad_norm": 0.5085616042672392,
+      "learning_rate": 2.8435457571870527e-06,
+      "loss": 0.1747,
+      "step": 2145
+    },
+    {
+      "epoch": 0.48351029374489535,
+      "grad_norm": 0.44676664208314826,
+      "learning_rate": 2.8417004362362465e-06,
+      "loss": 0.131,
+      "step": 2146
+    },
+    {
+      "epoch": 0.4837356014307038,
+      "grad_norm": 0.47361874739111104,
+      "learning_rate": 2.8398549255519237e-06,
+      "loss": 0.1602,
+      "step": 2147
+    },
+    {
+      "epoch": 0.48396090911651224,
+      "grad_norm": 0.4763300815360598,
+      "learning_rate": 2.838009226158829e-06,
+      "loss": 0.1634,
+      "step": 2148
+    },
+    {
+      "epoch": 0.48418621680232066,
+      "grad_norm": 0.5197625242097161,
+      "learning_rate": 2.83616333908181e-06,
+      "loss": 0.1731,
+      "step": 2149
+    },
+    {
+      "epoch": 0.4844115244881291,
+      "grad_norm": 0.4824667341017959,
+      "learning_rate": 2.8343172653458194e-06,
+      "loss": 0.1648,
+      "step": 2150
+    },
+    {
+      "epoch": 0.48463683217393755,
+      "grad_norm": 0.4719192705898006,
+      "learning_rate": 2.8324710059759126e-06,
+      "loss": 0.1591,
+      "step": 2151
+    },
+    {
+      "epoch": 0.48486213985974597,
+      "grad_norm": 0.4515406394716849,
+      "learning_rate": 2.8306245619972476e-06,
+      "loss": 0.1483,
+      "step": 2152
+    },
+    {
+      "epoch": 0.4850874475455544,
+      "grad_norm": 0.4818085870458921,
+      "learning_rate": 2.828777934435088e-06,
+      "loss": 0.1657,
+      "step": 2153
+    },
+    {
+      "epoch": 0.4853127552313628,
+      "grad_norm": 0.4948205614656545,
+      "learning_rate": 2.826931124314796e-06,
+      "loss": 0.1691,
+      "step": 2154
+    },
+    {
+      "epoch": 0.4855380629171713,
+      "grad_norm": 0.47291934012520137,
+      "learning_rate": 2.8250841326618367e-06,
+      "loss": 0.1527,
+      "step": 2155
+    },
+    {
+      "epoch": 0.4857633706029797,
+      "grad_norm": 0.48355294932750525,
+      "learning_rate": 2.8232369605017757e-06,
+      "loss": 0.1592,
+      "step": 2156
+    },
+    {
+      "epoch": 0.4859886782887881,
+      "grad_norm": 0.4768913578131576,
+      "learning_rate": 2.8213896088602786e-06,
+      "loss": 0.1487,
+      "step": 2157
+    },
+    {
+      "epoch": 0.4862139859745966,
+      "grad_norm": 0.5053824042202755,
+      "learning_rate": 2.8195420787631113e-06,
+      "loss": 0.1714,
+      "step": 2158
+    },
+    {
+      "epoch": 0.486439293660405,
+      "grad_norm": 0.487503834880126,
+      "learning_rate": 2.8176943712361394e-06,
+      "loss": 0.1627,
+      "step": 2159
+    },
+    {
+      "epoch": 0.4866646013462134,
+      "grad_norm": 0.4992652219189248,
+      "learning_rate": 2.8158464873053236e-06,
+      "loss": 0.153,
+      "step": 2160
+    },
+    {
+      "epoch": 0.48688990903202184,
+      "grad_norm": 0.4781211578280688,
+      "learning_rate": 2.8139984279967265e-06,
+      "loss": 0.1569,
+      "step": 2161
+    },
+    {
+      "epoch": 0.4871152167178303,
+      "grad_norm": 0.45967039527673825,
+      "learning_rate": 2.8121501943365066e-06,
+      "loss": 0.1404,
+      "step": 2162
+    },
+    {
+      "epoch": 0.48734052440363873,
+      "grad_norm": 0.46645982532415253,
+      "learning_rate": 2.810301787350918e-06,
+      "loss": 0.1562,
+      "step": 2163
+    },
+    {
+      "epoch": 0.48756583208944715,
+      "grad_norm": 0.46918371300837053,
+      "learning_rate": 2.808453208066314e-06,
+      "loss": 0.1465,
+      "step": 2164
+    },
+    {
+      "epoch": 0.48779113977525557,
+      "grad_norm": 0.46548524196093854,
+      "learning_rate": 2.8066044575091404e-06,
+      "loss": 0.1522,
+      "step": 2165
+    },
+    {
+      "epoch": 0.48801644746106404,
+      "grad_norm": 0.5289539976247626,
+      "learning_rate": 2.8047555367059404e-06,
+      "loss": 0.1739,
+      "step": 2166
+    },
+    {
+      "epoch": 0.48824175514687246,
+      "grad_norm": 0.458664098217161,
+      "learning_rate": 2.80290644668335e-06,
+      "loss": 0.1405,
+      "step": 2167
+    },
+    {
+      "epoch": 0.4884670628326809,
+      "grad_norm": 0.4717126138511026,
+      "learning_rate": 2.8010571884681004e-06,
+      "loss": 0.1542,
+      "step": 2168
+    },
+    {
+      "epoch": 0.4886923705184893,
+      "grad_norm": 0.4601707871424624,
+      "learning_rate": 2.799207763087015e-06,
+      "loss": 0.1484,
+      "step": 2169
+    },
+    {
+      "epoch": 0.48891767820429777,
+      "grad_norm": 0.4711813562824451,
+      "learning_rate": 2.7973581715670124e-06,
+      "loss": 0.1492,
+      "step": 2170
+    },
+    {
+      "epoch": 0.4891429858901062,
+      "grad_norm": 0.4512983700781856,
+      "learning_rate": 2.7955084149351002e-06,
+      "loss": 0.1559,
+      "step": 2171
+    },
+    {
+      "epoch": 0.4893682935759146,
+      "grad_norm": 0.49222156857332283,
+      "learning_rate": 2.7936584942183804e-06,
+      "loss": 0.1596,
+      "step": 2172
+    },
+    {
+      "epoch": 0.489593601261723,
+      "grad_norm": 0.46487770605974843,
+      "learning_rate": 2.7918084104440446e-06,
+      "loss": 0.1534,
+      "step": 2173
+    },
+    {
+      "epoch": 0.4898189089475315,
+      "grad_norm": 0.46155781463668333,
+      "learning_rate": 2.7899581646393746e-06,
+      "loss": 0.1585,
+      "step": 2174
+    },
+    {
+      "epoch": 0.4900442166333399,
+      "grad_norm": 0.48435801828172664,
+      "learning_rate": 2.7881077578317445e-06,
+      "loss": 0.1561,
+      "step": 2175
+    },
+    {
+      "epoch": 0.49026952431914833,
+      "grad_norm": 0.46865565885766813,
+      "learning_rate": 2.7862571910486148e-06,
+      "loss": 0.155,
+      "step": 2176
+    },
+    {
+      "epoch": 0.49049483200495675,
+      "grad_norm": 0.5237421496925503,
+      "learning_rate": 2.784406465317538e-06,
+      "loss": 0.1592,
+      "step": 2177
+    },
+    {
+      "epoch": 0.4907201396907652,
+      "grad_norm": 0.5063351392528739,
+      "learning_rate": 2.7825555816661503e-06,
+      "loss": 0.176,
+      "step": 2178
+    },
+    {
+      "epoch": 0.49094544737657364,
+      "grad_norm": 0.45029290239029635,
+      "learning_rate": 2.7807045411221813e-06,
+      "loss": 0.1497,
+      "step": 2179
+    },
+    {
+      "epoch": 0.49117075506238206,
+      "grad_norm": 0.4807262732955275,
+      "learning_rate": 2.778853344713443e-06,
+      "loss": 0.1523,
+      "step": 2180
+    },
+    {
+      "epoch": 0.4913960627481905,
+      "grad_norm": 0.4527240365909731,
+      "learning_rate": 2.777001993467837e-06,
+      "loss": 0.1503,
+      "step": 2181
+    },
+    {
+      "epoch": 0.49162137043399895,
+      "grad_norm": 0.4750132640988713,
+      "learning_rate": 2.7751504884133484e-06,
+      "loss": 0.1504,
+      "step": 2182
+    },
+    {
+      "epoch": 0.49184667811980737,
+      "grad_norm": 0.5064783537919353,
+      "learning_rate": 2.7732988305780496e-06,
+      "loss": 0.1604,
+      "step": 2183
+    },
+    {
+      "epoch": 0.4920719858056158,
+      "grad_norm": 0.4501987876305936,
+      "learning_rate": 2.7714470209900974e-06,
+      "loss": 0.1512,
+      "step": 2184
+    },
+    {
+      "epoch": 0.4922972934914242,
+      "grad_norm": 0.4760064527034139,
+      "learning_rate": 2.769595060677732e-06,
+      "loss": 0.149,
+      "step": 2185
+    },
+    {
+      "epoch": 0.4925226011772327,
+      "grad_norm": 0.5027689026168123,
+      "learning_rate": 2.7677429506692788e-06,
+      "loss": 0.1527,
+      "step": 2186
+    },
+    {
+      "epoch": 0.4927479088630411,
+      "grad_norm": 0.4722739354725667,
+      "learning_rate": 2.7658906919931443e-06,
+      "loss": 0.1512,
+      "step": 2187
+    },
+    {
+      "epoch": 0.4929732165488495,
+      "grad_norm": 0.48734288018027916,
+      "learning_rate": 2.76403828567782e-06,
+      "loss": 0.1553,
+      "step": 2188
+    },
+    {
+      "epoch": 0.49319852423465793,
+      "grad_norm": 0.46063676497770223,
+      "learning_rate": 2.7621857327518763e-06,
+      "loss": 0.1609,
+      "step": 2189
+    },
+    {
+      "epoch": 0.4934238319204664,
+      "grad_norm": 0.4881192023888224,
+      "learning_rate": 2.7603330342439686e-06,
+      "loss": 0.1637,
+      "step": 2190
+    },
+    {
+      "epoch": 0.4936491396062748,
+      "grad_norm": 0.46588329681809043,
+      "learning_rate": 2.7584801911828314e-06,
+      "loss": 0.1597,
+      "step": 2191
+    },
+    {
+      "epoch": 0.49387444729208324,
+      "grad_norm": 0.48151698956393363,
+      "learning_rate": 2.7566272045972777e-06,
+      "loss": 0.1555,
+      "step": 2192
+    },
+    {
+      "epoch": 0.49409975497789166,
+      "grad_norm": 0.44929106962011756,
+      "learning_rate": 2.7547740755162034e-06,
+      "loss": 0.139,
+      "step": 2193
+    },
+    {
+      "epoch": 0.49432506266370013,
+      "grad_norm": 0.4619821656904074,
+      "learning_rate": 2.752920804968581e-06,
+      "loss": 0.1516,
+      "step": 2194
+    },
+    {
+      "epoch": 0.49455037034950855,
+      "grad_norm": 0.4624932946392754,
+      "learning_rate": 2.7510673939834633e-06,
+      "loss": 0.1579,
+      "step": 2195
+    },
+    {
+      "epoch": 0.49477567803531697,
+      "grad_norm": 0.44059934991547617,
+      "learning_rate": 2.7492138435899794e-06,
+      "loss": 0.1533,
+      "step": 2196
+    },
+    {
+      "epoch": 0.4950009857211254,
+      "grad_norm": 0.46495736661665293,
+      "learning_rate": 2.747360154817338e-06,
+      "loss": 0.1568,
+      "step": 2197
+    },
+    {
+      "epoch": 0.49522629340693386,
+      "grad_norm": 0.4471263495684936,
+      "learning_rate": 2.745506328694822e-06,
+      "loss": 0.1505,
+      "step": 2198
+    },
+    {
+      "epoch": 0.4954516010927423,
+      "grad_norm": 0.47502698148207845,
+      "learning_rate": 2.743652366251793e-06,
+      "loss": 0.1528,
+      "step": 2199
+    },
+    {
+      "epoch": 0.4956769087785507,
+      "grad_norm": 0.474101185921597,
+      "learning_rate": 2.741798268517687e-06,
+      "loss": 0.1444,
+      "step": 2200
+    },
+    {
+      "epoch": 0.4959022164643591,
+      "grad_norm": 0.4950438275209808,
+      "learning_rate": 2.7399440365220153e-06,
+      "loss": 0.159,
+      "step": 2201
+    },
+    {
+      "epoch": 0.4961275241501676,
+      "grad_norm": 0.48514409432512545,
+      "learning_rate": 2.738089671294364e-06,
+      "loss": 0.1623,
+      "step": 2202
+    },
+    {
+      "epoch": 0.496352831835976,
+      "grad_norm": 0.49897046758102964,
+      "learning_rate": 2.7362351738643926e-06,
+      "loss": 0.1714,
+      "step": 2203
+    },
+    {
+      "epoch": 0.4965781395217844,
+      "grad_norm": 0.4944720313099016,
+      "learning_rate": 2.734380545261835e-06,
+      "loss": 0.1454,
+      "step": 2204
+    },
+    {
+      "epoch": 0.4968034472075929,
+      "grad_norm": 0.4740848361304964,
+      "learning_rate": 2.7325257865164955e-06,
+      "loss": 0.1505,
+      "step": 2205
+    },
+    {
+      "epoch": 0.4970287548934013,
+      "grad_norm": 0.5044552603154835,
+      "learning_rate": 2.730670898658255e-06,
+      "loss": 0.1779,
+      "step": 2206
+    },
+    {
+      "epoch": 0.49725406257920973,
+      "grad_norm": 0.46394146351071547,
+      "learning_rate": 2.7288158827170623e-06,
+      "loss": 0.1396,
+      "step": 2207
+    },
+    {
+      "epoch": 0.49747937026501815,
+      "grad_norm": 0.49280257361880014,
+      "learning_rate": 2.726960739722939e-06,
+      "loss": 0.1622,
+      "step": 2208
+    },
+    {
+      "epoch": 0.4977046779508266,
+      "grad_norm": 0.4867686438552059,
+      "learning_rate": 2.725105470705977e-06,
+      "loss": 0.1581,
+      "step": 2209
+    },
+    {
+      "epoch": 0.49792998563663504,
+      "grad_norm": 0.5219938141413315,
+      "learning_rate": 2.7232500766963373e-06,
+      "loss": 0.1686,
+      "step": 2210
+    },
+    {
+      "epoch": 0.49815529332244346,
+      "grad_norm": 0.4764321227719793,
+      "learning_rate": 2.7213945587242507e-06,
+      "loss": 0.1567,
+      "step": 2211
+    },
+    {
+      "epoch": 0.4983806010082519,
+      "grad_norm": 0.49384595085813865,
+      "learning_rate": 2.7195389178200194e-06,
+      "loss": 0.1583,
+      "step": 2212
+    },
+    {
+      "epoch": 0.49860590869406035,
+      "grad_norm": 0.5075757200906527,
+      "learning_rate": 2.7176831550140093e-06,
+      "loss": 0.1676,
+      "step": 2213
+    },
+    {
+      "epoch": 0.49883121637986877,
+      "grad_norm": 0.4524722468045472,
+      "learning_rate": 2.7158272713366573e-06,
+      "loss": 0.1384,
+      "step": 2214
+    },
+    {
+      "epoch": 0.4990565240656772,
+      "grad_norm": 0.5083060899486224,
+      "learning_rate": 2.713971267818466e-06,
+      "loss": 0.1613,
+      "step": 2215
+    },
+    {
+      "epoch": 0.4992818317514856,
+      "grad_norm": 0.4396851281617713,
+      "learning_rate": 2.7121151454900048e-06,
+      "loss": 0.1482,
+      "step": 2216
+    },
+    {
+      "epoch": 0.4995071394372941,
+      "grad_norm": 0.46962633997335007,
+      "learning_rate": 2.7102589053819107e-06,
+      "loss": 0.147,
+      "step": 2217
+    },
+    {
+      "epoch": 0.4997324471231025,
+      "grad_norm": 0.5126065467533945,
+      "learning_rate": 2.7084025485248827e-06,
+      "loss": 0.1575,
+      "step": 2218
+    },
+    {
+      "epoch": 0.4999577548089109,
+      "grad_norm": 0.4636702745429642,
+      "learning_rate": 2.706546075949688e-06,
+      "loss": 0.1497,
+      "step": 2219
+    },
+    {
+      "epoch": 0.5001830624947193,
+      "grad_norm": 0.4624702418642572,
+      "learning_rate": 2.7046894886871564e-06,
+      "loss": 0.1554,
+      "step": 2220
+    },
+    {
+      "epoch": 0.5004083701805278,
+      "grad_norm": 0.4919207689507557,
+      "learning_rate": 2.7028327877681808e-06,
+      "loss": 0.1567,
+      "step": 2221
+    },
+    {
+      "epoch": 0.5006336778663362,
+      "grad_norm": 0.5157069877586713,
+      "learning_rate": 2.700975974223719e-06,
+      "loss": 0.1561,
+      "step": 2222
+    },
+    {
+      "epoch": 0.5008589855521447,
+      "grad_norm": 0.45397574648979233,
+      "learning_rate": 2.6991190490847898e-06,
+      "loss": 0.1463,
+      "step": 2223
+    },
+    {
+      "epoch": 0.5010842932379531,
+      "grad_norm": 0.4805592141525948,
+      "learning_rate": 2.6972620133824745e-06,
+      "loss": 0.1544,
+      "step": 2224
+    },
+    {
+      "epoch": 0.5013096009237615,
+      "grad_norm": 0.45753171981042684,
+      "learning_rate": 2.695404868147916e-06,
+      "loss": 0.1521,
+      "step": 2225
+    },
+    {
+      "epoch": 0.50153490860957,
+      "grad_norm": 0.4763708134097829,
+      "learning_rate": 2.6935476144123173e-06,
+      "loss": 0.1489,
+      "step": 2226
+    },
+    {
+      "epoch": 0.5017602162953784,
+      "grad_norm": 0.4753257879884957,
+      "learning_rate": 2.691690253206943e-06,
+      "loss": 0.1556,
+      "step": 2227
+    },
+    {
+      "epoch": 0.5019855239811868,
+      "grad_norm": 0.4927561800774899,
+      "learning_rate": 2.689832785563116e-06,
+      "loss": 0.1467,
+      "step": 2228
+    },
+    {
+      "epoch": 0.5022108316669952,
+      "grad_norm": 0.4890552682019787,
+      "learning_rate": 2.6879752125122193e-06,
+      "loss": 0.1476,
+      "step": 2229
+    },
+    {
+      "epoch": 0.5024361393528036,
+      "grad_norm": 0.5019953873414026,
+      "learning_rate": 2.6861175350856937e-06,
+      "loss": 0.1615,
+      "step": 2230
+    },
+    {
+      "epoch": 0.5026614470386122,
+      "grad_norm": 0.5010903995101,
+      "learning_rate": 2.684259754315038e-06,
+      "loss": 0.1736,
+      "step": 2231
+    },
+    {
+      "epoch": 0.5028867547244206,
+      "grad_norm": 0.5022799823142762,
+      "learning_rate": 2.6824018712318084e-06,
+      "loss": 0.154,
+      "step": 2232
+    },
+    {
+      "epoch": 0.503112062410229,
+      "grad_norm": 0.5222667007278232,
+      "learning_rate": 2.6805438868676186e-06,
+      "loss": 0.1574,
+      "step": 2233
+    },
+    {
+      "epoch": 0.5033373700960374,
+      "grad_norm": 0.4734865357653425,
+      "learning_rate": 2.6786858022541385e-06,
+      "loss": 0.1487,
+      "step": 2234
+    },
+    {
+      "epoch": 0.5035626777818458,
+      "grad_norm": 0.49629092969775523,
+      "learning_rate": 2.676827618423093e-06,
+      "loss": 0.169,
+      "step": 2235
+    },
+    {
+      "epoch": 0.5037879854676542,
+      "grad_norm": 0.47744881578341003,
+      "learning_rate": 2.674969336406262e-06,
+      "loss": 0.1543,
+      "step": 2236
+    },
+    {
+      "epoch": 0.5040132931534627,
+      "grad_norm": 0.49005111881485025,
+      "learning_rate": 2.6731109572354795e-06,
+      "loss": 0.1615,
+      "step": 2237
+    },
+    {
+      "epoch": 0.5042386008392711,
+      "grad_norm": 0.5405479885895289,
+      "learning_rate": 2.6712524819426355e-06,
+      "loss": 0.1741,
+      "step": 2238
+    },
+    {
+      "epoch": 0.5044639085250796,
+      "grad_norm": 0.5018883042516251,
+      "learning_rate": 2.6693939115596718e-06,
+      "loss": 0.1542,
+      "step": 2239
+    },
+    {
+      "epoch": 0.504689216210888,
+      "grad_norm": 0.44832772942283,
+      "learning_rate": 2.6675352471185824e-06,
+      "loss": 0.1432,
+      "step": 2240
+    },
+    {
+      "epoch": 0.5049145238966964,
+      "grad_norm": 0.47367279205265483,
+      "learning_rate": 2.6656764896514152e-06,
+      "loss": 0.1511,
+      "step": 2241
+    },
+    {
+      "epoch": 0.5051398315825049,
+      "grad_norm": 0.47569526383637317,
+      "learning_rate": 2.6638176401902693e-06,
+      "loss": 0.1639,
+      "step": 2242
+    },
+    {
+      "epoch": 0.5053651392683133,
+      "grad_norm": 0.4650884586981442,
+      "learning_rate": 2.6619586997672923e-06,
+      "loss": 0.1429,
+      "step": 2243
+    },
+    {
+      "epoch": 0.5055904469541217,
+      "grad_norm": 0.4823395176993751,
+      "learning_rate": 2.6600996694146876e-06,
+      "loss": 0.1668,
+      "step": 2244
+    },
+    {
+      "epoch": 0.5058157546399301,
+      "grad_norm": 0.43899046677750514,
+      "learning_rate": 2.658240550164704e-06,
+      "loss": 0.1468,
+      "step": 2245
+    },
+    {
+      "epoch": 0.5060410623257385,
+      "grad_norm": 0.462906398923942,
+      "learning_rate": 2.656381343049641e-06,
+      "loss": 0.1509,
+      "step": 2246
+    },
+    {
+      "epoch": 0.5062663700115471,
+      "grad_norm": 0.46216886541797236,
+      "learning_rate": 2.654522049101847e-06,
+      "loss": 0.1454,
+      "step": 2247
+    },
+    {
+      "epoch": 0.5064916776973555,
+      "grad_norm": 0.47781029672794806,
+      "learning_rate": 2.652662669353719e-06,
+      "loss": 0.1625,
+      "step": 2248
+    },
+    {
+      "epoch": 0.5067169853831639,
+      "grad_norm": 0.4687429076107305,
+      "learning_rate": 2.6508032048377006e-06,
+      "loss": 0.138,
+      "step": 2249
+    },
+    {
+      "epoch": 0.5069422930689723,
+      "grad_norm": 0.49544551469790465,
+      "learning_rate": 2.648943656586284e-06,
+      "loss": 0.1473,
+      "step": 2250
+    },
+    {
+      "epoch": 0.5071676007547807,
+      "grad_norm": 0.47331775745367966,
+      "learning_rate": 2.6470840256320064e-06,
+      "loss": 0.1524,
+      "step": 2251
+    },
+    {
+      "epoch": 0.5073929084405892,
+      "grad_norm": 0.423740267378944,
+      "learning_rate": 2.6452243130074523e-06,
+      "loss": 0.1283,
+      "step": 2252
+    },
+    {
+      "epoch": 0.5076182161263976,
+      "grad_norm": 0.46962236211799213,
+      "learning_rate": 2.6433645197452493e-06,
+      "loss": 0.1499,
+      "step": 2253
+    },
+    {
+      "epoch": 0.5078435238122061,
+      "grad_norm": 0.4931593528990676,
+      "learning_rate": 2.6415046468780726e-06,
+      "loss": 0.1455,
+      "step": 2254
+    },
+    {
+      "epoch": 0.5080688314980145,
+      "grad_norm": 0.4633383428887458,
+      "learning_rate": 2.63964469543864e-06,
+      "loss": 0.1384,
+      "step": 2255
+    },
+    {
+      "epoch": 0.5082941391838229,
+      "grad_norm": 0.5014476308647303,
+      "learning_rate": 2.637784666459714e-06,
+      "loss": 0.1525,
+      "step": 2256
+    },
+    {
+      "epoch": 0.5085194468696314,
+      "grad_norm": 0.5187286528852701,
+      "learning_rate": 2.635924560974098e-06,
+      "loss": 0.1672,
+      "step": 2257
+    },
+    {
+      "epoch": 0.5087447545554398,
+      "grad_norm": 0.48859410138629195,
+      "learning_rate": 2.6340643800146387e-06,
+      "loss": 0.1577,
+      "step": 2258
+    },
+    {
+      "epoch": 0.5089700622412482,
+      "grad_norm": 0.4758161336287191,
+      "learning_rate": 2.6322041246142273e-06,
+      "loss": 0.1554,
+      "step": 2259
+    },
+    {
+      "epoch": 0.5091953699270566,
+      "grad_norm": 0.5204038587103618,
+      "learning_rate": 2.6303437958057932e-06,
+      "loss": 0.1596,
+      "step": 2260
+    },
+    {
+      "epoch": 0.509420677612865,
+      "grad_norm": 0.4755339689407434,
+      "learning_rate": 2.6284833946223075e-06,
+      "loss": 0.1514,
+      "step": 2261
+    },
+    {
+      "epoch": 0.5096459852986736,
+      "grad_norm": 0.49529525973555916,
+      "learning_rate": 2.626622922096782e-06,
+      "loss": 0.1662,
+      "step": 2262
+    },
+    {
+      "epoch": 0.509871292984482,
+      "grad_norm": 0.47102844584098646,
+      "learning_rate": 2.624762379262268e-06,
+      "loss": 0.1497,
+      "step": 2263
+    },
+    {
+      "epoch": 0.5100966006702904,
+      "grad_norm": 0.4819251906829047,
+      "learning_rate": 2.622901767151855e-06,
+      "loss": 0.1532,
+      "step": 2264
+    },
+    {
+      "epoch": 0.5103219083560988,
+      "grad_norm": 0.472069862823219,
+      "learning_rate": 2.6210410867986713e-06,
+      "loss": 0.1436,
+      "step": 2265
+    },
+    {
+      "epoch": 0.5105472160419072,
+      "grad_norm": 0.4645494432026978,
+      "learning_rate": 2.619180339235884e-06,
+      "loss": 0.1422,
+      "step": 2266
+    },
+    {
+      "epoch": 0.5107725237277156,
+      "grad_norm": 0.47263922852142376,
+      "learning_rate": 2.6173195254966966e-06,
+      "loss": 0.162,
+      "step": 2267
+    },
+    {
+      "epoch": 0.5109978314135241,
+      "grad_norm": 0.46779769487146516,
+      "learning_rate": 2.6154586466143495e-06,
+      "loss": 0.143,
+      "step": 2268
+    },
+    {
+      "epoch": 0.5112231390993325,
+      "grad_norm": 0.4600174066739792,
+      "learning_rate": 2.6135977036221195e-06,
+      "loss": 0.1442,
+      "step": 2269
+    },
+    {
+      "epoch": 0.511448446785141,
+      "grad_norm": 0.4575400490189286,
+      "learning_rate": 2.6117366975533187e-06,
+      "loss": 0.1512,
+      "step": 2270
+    },
+    {
+      "epoch": 0.5116737544709494,
+      "grad_norm": 0.4826461966104453,
+      "learning_rate": 2.609875629441295e-06,
+      "loss": 0.1638,
+      "step": 2271
+    },
+    {
+      "epoch": 0.5118990621567578,
+      "grad_norm": 0.4707429777697515,
+      "learning_rate": 2.60801450031943e-06,
+      "loss": 0.1469,
+      "step": 2272
+    },
+    {
+      "epoch": 0.5121243698425663,
+      "grad_norm": 0.46468370484946786,
+      "learning_rate": 2.6061533112211394e-06,
+      "loss": 0.1532,
+      "step": 2273
+    },
+    {
+      "epoch": 0.5123496775283747,
+      "grad_norm": 0.49977089530261853,
+      "learning_rate": 2.604292063179871e-06,
+      "loss": 0.1603,
+      "step": 2274
+    },
+    {
+      "epoch": 0.5125749852141831,
+      "grad_norm": 0.4840088481175901,
+      "learning_rate": 2.602430757229108e-06,
+      "loss": 0.1599,
+      "step": 2275
+    },
+    {
+      "epoch": 0.5128002928999915,
+      "grad_norm": 0.48088896715216267,
+      "learning_rate": 2.600569394402363e-06,
+      "loss": 0.1533,
+      "step": 2276
+    },
+    {
+      "epoch": 0.5130256005857999,
+      "grad_norm": 0.5062988229080415,
+      "learning_rate": 2.5987079757331824e-06,
+      "loss": 0.1666,
+      "step": 2277
+    },
+    {
+      "epoch": 0.5132509082716085,
+      "grad_norm": 0.4991590684305053,
+      "learning_rate": 2.596846502255142e-06,
+      "loss": 0.1495,
+      "step": 2278
+    },
+    {
+      "epoch": 0.5134762159574169,
+      "grad_norm": 0.4712901894078783,
+      "learning_rate": 2.5949849750018486e-06,
+      "loss": 0.1494,
+      "step": 2279
+    },
+    {
+      "epoch": 0.5137015236432253,
+      "grad_norm": 0.4566845858422049,
+      "learning_rate": 2.5931233950069385e-06,
+      "loss": 0.1481,
+      "step": 2280
+    },
+    {
+      "epoch": 0.5139268313290337,
+      "grad_norm": 0.468692135600313,
+      "learning_rate": 2.591261763304079e-06,
+      "loss": 0.152,
+      "step": 2281
+    },
+    {
+      "epoch": 0.5141521390148421,
+      "grad_norm": 0.48731655698950843,
+      "learning_rate": 2.589400080926964e-06,
+      "loss": 0.1469,
+      "step": 2282
+    },
+    {
+      "epoch": 0.5143774467006506,
+      "grad_norm": 0.47987896655282924,
+      "learning_rate": 2.5875383489093165e-06,
+      "loss": 0.1613,
+      "step": 2283
+    },
+    {
+      "epoch": 0.514602754386459,
+      "grad_norm": 0.4635204806630267,
+      "learning_rate": 2.585676568284886e-06,
+      "loss": 0.1532,
+      "step": 2284
+    },
+    {
+      "epoch": 0.5148280620722674,
+      "grad_norm": 0.47497923474665377,
+      "learning_rate": 2.583814740087451e-06,
+      "loss": 0.1462,
+      "step": 2285
+    },
+    {
+      "epoch": 0.5150533697580759,
+      "grad_norm": 0.5202422022869216,
+      "learning_rate": 2.581952865350815e-06,
+      "loss": 0.1636,
+      "step": 2286
+    },
+    {
+      "epoch": 0.5152786774438843,
+      "grad_norm": 0.46226673508432603,
+      "learning_rate": 2.5800909451088075e-06,
+      "loss": 0.1558,
+      "step": 2287
+    },
+    {
+      "epoch": 0.5155039851296928,
+      "grad_norm": 0.484793426628321,
+      "learning_rate": 2.578228980395283e-06,
+      "loss": 0.1523,
+      "step": 2288
+    },
+    {
+      "epoch": 0.5157292928155012,
+      "grad_norm": 0.49786647707597725,
+      "learning_rate": 2.5763669722441226e-06,
+      "loss": 0.1626,
+      "step": 2289
+    },
+    {
+      "epoch": 0.5159546005013096,
+      "grad_norm": 0.44749134037384025,
+      "learning_rate": 2.5745049216892286e-06,
+      "loss": 0.1376,
+      "step": 2290
+    },
+    {
+      "epoch": 0.516179908187118,
+      "grad_norm": 0.4943654328069877,
+      "learning_rate": 2.5726428297645285e-06,
+      "loss": 0.1504,
+      "step": 2291
+    },
+    {
+      "epoch": 0.5164052158729264,
+      "grad_norm": 0.4962431635915529,
+      "learning_rate": 2.570780697503973e-06,
+      "loss": 0.1528,
+      "step": 2292
+    },
+    {
+      "epoch": 0.5166305235587348,
+      "grad_norm": 0.47973962504827283,
+      "learning_rate": 2.5689185259415346e-06,
+      "loss": 0.1676,
+      "step": 2293
+    },
+    {
+      "epoch": 0.5168558312445434,
+      "grad_norm": 0.4546694046351566,
+      "learning_rate": 2.5670563161112073e-06,
+      "loss": 0.1419,
+      "step": 2294
+    },
+    {
+      "epoch": 0.5170811389303518,
+      "grad_norm": 0.5004005389250398,
+      "learning_rate": 2.5651940690470074e-06,
+      "loss": 0.1578,
+      "step": 2295
+    },
+    {
+      "epoch": 0.5173064466161602,
+      "grad_norm": 0.5551306967913575,
+      "learning_rate": 2.56333178578297e-06,
+      "loss": 0.1632,
+      "step": 2296
+    },
+    {
+      "epoch": 0.5175317543019686,
+      "grad_norm": 0.4957720566111589,
+      "learning_rate": 2.5614694673531533e-06,
+      "loss": 0.1515,
+      "step": 2297
+    },
+    {
+      "epoch": 0.517757061987777,
+      "grad_norm": 0.4733168651798975,
+      "learning_rate": 2.5596071147916325e-06,
+      "loss": 0.1433,
+      "step": 2298
+    },
+    {
+      "epoch": 0.5179823696735855,
+      "grad_norm": 0.4984969498794251,
+      "learning_rate": 2.557744729132503e-06,
+      "loss": 0.168,
+      "step": 2299
+    },
+    {
+      "epoch": 0.5182076773593939,
+      "grad_norm": 0.4739201510380568,
+      "learning_rate": 2.555882311409878e-06,
+      "loss": 0.1616,
+      "step": 2300
+    },
+    {
+      "epoch": 0.5184329850452023,
+      "grad_norm": 0.47477837369247494,
+      "learning_rate": 2.554019862657888e-06,
+      "loss": 0.1512,
+      "step": 2301
+    },
+    {
+      "epoch": 0.5186582927310108,
+      "grad_norm": 0.4827657739269289,
+      "learning_rate": 2.5521573839106815e-06,
+      "loss": 0.1532,
+      "step": 2302
+    },
+    {
+      "epoch": 0.5188836004168192,
+      "grad_norm": 0.48081583349713264,
+      "learning_rate": 2.5502948762024244e-06,
+      "loss": 0.1487,
+      "step": 2303
+    },
+    {
+      "epoch": 0.5191089081026277,
+      "grad_norm": 0.5136945737827481,
+      "learning_rate": 2.5484323405672965e-06,
+      "loss": 0.163,
+      "step": 2304
+    },
+    {
+      "epoch": 0.5193342157884361,
+      "grad_norm": 0.4852412918348156,
+      "learning_rate": 2.546569778039496e-06,
+      "loss": 0.1502,
+      "step": 2305
+    },
+    {
+      "epoch": 0.5195595234742445,
+      "grad_norm": 0.4839387174700844,
+      "learning_rate": 2.544707189653233e-06,
+      "loss": 0.1566,
+      "step": 2306
+    },
+    {
+      "epoch": 0.5197848311600529,
+      "grad_norm": 0.47852121774050516,
+      "learning_rate": 2.542844576442734e-06,
+      "loss": 0.1562,
+      "step": 2307
+    },
+    {
+      "epoch": 0.5200101388458613,
+      "grad_norm": 0.48176737230647965,
+      "learning_rate": 2.5409819394422386e-06,
+      "loss": 0.1473,
+      "step": 2308
+    },
+    {
+      "epoch": 0.5202354465316699,
+      "grad_norm": 0.4493722340870022,
+      "learning_rate": 2.539119279686001e-06,
+      "loss": 0.1389,
+      "step": 2309
+    },
+    {
+      "epoch": 0.5204607542174783,
+      "grad_norm": 0.4867659173267954,
+      "learning_rate": 2.5372565982082843e-06,
+      "loss": 0.1565,
+      "step": 2310
+    },
+    {
+      "epoch": 0.5206860619032867,
+      "grad_norm": 0.49083074439740926,
+      "learning_rate": 2.535393896043368e-06,
+      "loss": 0.1693,
+      "step": 2311
+    },
+    {
+      "epoch": 0.5209113695890951,
+      "grad_norm": 0.5118582010278282,
+      "learning_rate": 2.5335311742255392e-06,
+      "loss": 0.165,
+      "step": 2312
+    },
+    {
+      "epoch": 0.5211366772749035,
+      "grad_norm": 0.46165246926341413,
+      "learning_rate": 2.5316684337891005e-06,
+      "loss": 0.1513,
+      "step": 2313
+    },
+    {
+      "epoch": 0.521361984960712,
+      "grad_norm": 0.5081771222693094,
+      "learning_rate": 2.5298056757683604e-06,
+      "loss": 0.1744,
+      "step": 2314
+    },
+    {
+      "epoch": 0.5215872926465204,
+      "grad_norm": 0.47259517752258834,
+      "learning_rate": 2.52794290119764e-06,
+      "loss": 0.1529,
+      "step": 2315
+    },
+    {
+      "epoch": 0.5218126003323288,
+      "grad_norm": 0.5412140292393451,
+      "learning_rate": 2.5260801111112677e-06,
+      "loss": 0.1769,
+      "step": 2316
+    },
+    {
+      "epoch": 0.5220379080181373,
+      "grad_norm": 0.46283220054109886,
+      "learning_rate": 2.5242173065435815e-06,
+      "loss": 0.1458,
+      "step": 2317
+    },
+    {
+      "epoch": 0.5222632157039457,
+      "grad_norm": 0.44276183424887405,
+      "learning_rate": 2.5223544885289287e-06,
+      "loss": 0.1353,
+      "step": 2318
+    },
+    {
+      "epoch": 0.5224885233897542,
+      "grad_norm": 0.4848664491076183,
+      "learning_rate": 2.5204916581016608e-06,
+      "loss": 0.1502,
+      "step": 2319
+    },
+    {
+      "epoch": 0.5227138310755626,
+      "grad_norm": 0.48383050929686455,
+      "learning_rate": 2.518628816296139e-06,
+      "loss": 0.1559,
+      "step": 2320
+    },
+    {
+      "epoch": 0.522939138761371,
+      "grad_norm": 0.47148303411187387,
+      "learning_rate": 2.5167659641467302e-06,
+      "loss": 0.1492,
+      "step": 2321
+    },
+    {
+      "epoch": 0.5231644464471794,
+      "grad_norm": 0.5130011837904479,
+      "learning_rate": 2.5149031026878063e-06,
+      "loss": 0.1543,
+      "step": 2322
+    },
+    {
+      "epoch": 0.5233897541329878,
+      "grad_norm": 0.43672925822618086,
+      "learning_rate": 2.5130402329537444e-06,
+      "loss": 0.1276,
+      "step": 2323
+    },
+    {
+      "epoch": 0.5236150618187962,
+      "grad_norm": 0.5096482667519452,
+      "learning_rate": 2.5111773559789277e-06,
+      "loss": 0.1668,
+      "step": 2324
+    },
+    {
+      "epoch": 0.5238403695046048,
+      "grad_norm": 0.4790070945167404,
+      "learning_rate": 2.509314472797742e-06,
+      "loss": 0.1481,
+      "step": 2325
+    },
+    {
+      "epoch": 0.5240656771904132,
+      "grad_norm": 0.4576983330267588,
+      "learning_rate": 2.5074515844445774e-06,
+      "loss": 0.1439,
+      "step": 2326
+    },
+    {
+      "epoch": 0.5242909848762216,
+      "grad_norm": 0.5058847652244145,
+      "learning_rate": 2.5055886919538247e-06,
+      "loss": 0.1759,
+      "step": 2327
+    },
+    {
+      "epoch": 0.52451629256203,
+      "grad_norm": 0.4661076795855495,
+      "learning_rate": 2.50372579635988e-06,
+      "loss": 0.1502,
+      "step": 2328
+    },
+    {
+      "epoch": 0.5247416002478384,
+      "grad_norm": 0.4492691768597872,
+      "learning_rate": 2.5018628986971395e-06,
+      "loss": 0.1341,
+      "step": 2329
+    },
+    {
+      "epoch": 0.5249669079336469,
+      "grad_norm": 0.4715357476203562,
+      "learning_rate": 2.5e-06,
+      "loss": 0.1458,
+      "step": 2330
+    },
+    {
+      "epoch": 0.5251922156194553,
+      "grad_norm": 0.5043033667691116,
+      "learning_rate": 2.4981371013028618e-06,
+      "loss": 0.1699,
+      "step": 2331
+    },
+    {
+      "epoch": 0.5254175233052637,
+      "grad_norm": 0.46576266805778144,
+      "learning_rate": 2.4962742036401213e-06,
+      "loss": 0.1467,
+      "step": 2332
+    },
+    {
+      "epoch": 0.5256428309910722,
+      "grad_norm": 0.47223610796977544,
+      "learning_rate": 2.494411308046176e-06,
+      "loss": 0.1458,
+      "step": 2333
+    },
+    {
+      "epoch": 0.5258681386768806,
+      "grad_norm": 0.4994925272719538,
+      "learning_rate": 2.4925484155554235e-06,
+      "loss": 0.1632,
+      "step": 2334
+    },
+    {
+      "epoch": 0.5260934463626891,
+      "grad_norm": 0.4905595343755695,
+      "learning_rate": 2.490685527202258e-06,
+      "loss": 0.173,
+      "step": 2335
+    },
+    {
+      "epoch": 0.5263187540484975,
+      "grad_norm": 0.47089124227730406,
+      "learning_rate": 2.4888226440210723e-06,
+      "loss": 0.147,
+      "step": 2336
+    },
+    {
+      "epoch": 0.5265440617343059,
+      "grad_norm": 0.4599527059597026,
+      "learning_rate": 2.4869597670462555e-06,
+      "loss": 0.1524,
+      "step": 2337
+    },
+    {
+      "epoch": 0.5267693694201143,
+      "grad_norm": 0.4808986795245479,
+      "learning_rate": 2.4850968973121945e-06,
+      "loss": 0.1634,
+      "step": 2338
+    },
+    {
+      "epoch": 0.5269946771059227,
+      "grad_norm": 0.48589161401299336,
+      "learning_rate": 2.483234035853271e-06,
+      "loss": 0.1572,
+      "step": 2339
+    },
+    {
+      "epoch": 0.5272199847917312,
+      "grad_norm": 0.47488130307449883,
+      "learning_rate": 2.481371183703862e-06,
+      "loss": 0.1421,
+      "step": 2340
+    },
+    {
+      "epoch": 0.5274452924775397,
+      "grad_norm": 0.47126856103360054,
+      "learning_rate": 2.4795083418983405e-06,
+      "loss": 0.159,
+      "step": 2341
+    },
+    {
+      "epoch": 0.5276706001633481,
+      "grad_norm": 0.4926033575294693,
+      "learning_rate": 2.477645511471073e-06,
+      "loss": 0.1596,
+      "step": 2342
+    },
+    {
+      "epoch": 0.5278959078491565,
+      "grad_norm": 0.46332393911961006,
+      "learning_rate": 2.475782693456419e-06,
+      "loss": 0.1378,
+      "step": 2343
+    },
+    {
+      "epoch": 0.5281212155349649,
+      "grad_norm": 0.4725650364239991,
+      "learning_rate": 2.473919888888733e-06,
+      "loss": 0.1363,
+      "step": 2344
+    },
+    {
+      "epoch": 0.5283465232207734,
+      "grad_norm": 0.502604824086372,
+      "learning_rate": 2.472057098802361e-06,
+      "loss": 0.1547,
+      "step": 2345
+    },
+    {
+      "epoch": 0.5285718309065818,
+      "grad_norm": 0.46349203187037064,
+      "learning_rate": 2.4701943242316405e-06,
+      "loss": 0.1404,
+      "step": 2346
+    },
+    {
+      "epoch": 0.5287971385923902,
+      "grad_norm": 0.4973121971815668,
+      "learning_rate": 2.4683315662109003e-06,
+      "loss": 0.1672,
+      "step": 2347
+    },
+    {
+      "epoch": 0.5290224462781986,
+      "grad_norm": 0.43225256940399115,
+      "learning_rate": 2.466468825774461e-06,
+      "loss": 0.1356,
+      "step": 2348
+    },
+    {
+      "epoch": 0.5292477539640071,
+      "grad_norm": 0.477633940104711,
+      "learning_rate": 2.464606103956633e-06,
+      "loss": 0.1485,
+      "step": 2349
+    },
+    {
+      "epoch": 0.5294730616498156,
+      "grad_norm": 0.49014865184145545,
+      "learning_rate": 2.462743401791716e-06,
+      "loss": 0.1621,
+      "step": 2350
+    },
+    {
+      "epoch": 0.529698369335624,
+      "grad_norm": 0.4980263724685209,
+      "learning_rate": 2.460880720314e-06,
+      "loss": 0.1558,
+      "step": 2351
+    },
+    {
+      "epoch": 0.5299236770214324,
+      "grad_norm": 0.49992308145356856,
+      "learning_rate": 2.4590180605577614e-06,
+      "loss": 0.155,
+      "step": 2352
+    },
+    {
+      "epoch": 0.5301489847072408,
+      "grad_norm": 0.47557074308194397,
+      "learning_rate": 2.4571554235572665e-06,
+      "loss": 0.1527,
+      "step": 2353
+    },
+    {
+      "epoch": 0.5303742923930492,
+      "grad_norm": 0.4686763140918017,
+      "learning_rate": 2.4552928103467677e-06,
+      "loss": 0.1545,
+      "step": 2354
+    },
+    {
+      "epoch": 0.5305996000788576,
+      "grad_norm": 0.49061218416270386,
+      "learning_rate": 2.4534302219605044e-06,
+      "loss": 0.1525,
+      "step": 2355
+    },
+    {
+      "epoch": 0.5308249077646662,
+      "grad_norm": 0.4635700122759651,
+      "learning_rate": 2.4515676594327035e-06,
+      "loss": 0.1452,
+      "step": 2356
+    },
+    {
+      "epoch": 0.5310502154504746,
+      "grad_norm": 0.48378534785787525,
+      "learning_rate": 2.4497051237975773e-06,
+      "loss": 0.1544,
+      "step": 2357
+    },
+    {
+      "epoch": 0.531275523136283,
+      "grad_norm": 0.504499518788173,
+      "learning_rate": 2.4478426160893197e-06,
+      "loss": 0.1666,
+      "step": 2358
+    },
+    {
+      "epoch": 0.5315008308220914,
+      "grad_norm": 0.4972981132140918,
+      "learning_rate": 2.4459801373421134e-06,
+      "loss": 0.1533,
+      "step": 2359
+    },
+    {
+      "epoch": 0.5317261385078998,
+      "grad_norm": 0.4770186193899861,
+      "learning_rate": 2.4441176885901234e-06,
+      "loss": 0.1524,
+      "step": 2360
+    },
+    {
+      "epoch": 0.5319514461937083,
+      "grad_norm": 0.48734897113858827,
+      "learning_rate": 2.4422552708674977e-06,
+      "loss": 0.1476,
+      "step": 2361
+    },
+    {
+      "epoch": 0.5321767538795167,
+      "grad_norm": 0.4816980687725706,
+      "learning_rate": 2.440392885208368e-06,
+      "loss": 0.143,
+      "step": 2362
+    },
+    {
+      "epoch": 0.5324020615653251,
+      "grad_norm": 0.47347405015746924,
+      "learning_rate": 2.4385305326468475e-06,
+      "loss": 0.1487,
+      "step": 2363
+    },
+    {
+      "epoch": 0.5326273692511336,
+      "grad_norm": 0.45637701573209427,
+      "learning_rate": 2.436668214217031e-06,
+      "loss": 0.1427,
+      "step": 2364
+    },
+    {
+      "epoch": 0.532852676936942,
+      "grad_norm": 0.46057914273271494,
+      "learning_rate": 2.4348059309529935e-06,
+      "loss": 0.1519,
+      "step": 2365
+    },
+    {
+      "epoch": 0.5330779846227505,
+      "grad_norm": 0.4618776570707972,
+      "learning_rate": 2.4329436838887936e-06,
+      "loss": 0.1455,
+      "step": 2366
+    },
+    {
+      "epoch": 0.5333032923085589,
+      "grad_norm": 0.5031610167097199,
+      "learning_rate": 2.4310814740584663e-06,
+      "loss": 0.1587,
+      "step": 2367
+    },
+    {
+      "epoch": 0.5335285999943673,
+      "grad_norm": 0.4851364088493003,
+      "learning_rate": 2.4292193024960275e-06,
+      "loss": 0.1534,
+      "step": 2368
+    },
+    {
+      "epoch": 0.5337539076801757,
+      "grad_norm": 0.44650754269598664,
+      "learning_rate": 2.427357170235472e-06,
+      "loss": 0.132,
+      "step": 2369
+    },
+    {
+      "epoch": 0.5339792153659841,
+      "grad_norm": 0.4834717917296514,
+      "learning_rate": 2.425495078310772e-06,
+      "loss": 0.1496,
+      "step": 2370
+    },
+    {
+      "epoch": 0.5342045230517926,
+      "grad_norm": 0.494643450442732,
+      "learning_rate": 2.4236330277558774e-06,
+      "loss": 0.1632,
+      "step": 2371
+    },
+    {
+      "epoch": 0.5344298307376011,
+      "grad_norm": 0.4847535814939488,
+      "learning_rate": 2.4217710196047166e-06,
+      "loss": 0.1609,
+      "step": 2372
+    },
+    {
+      "epoch": 0.5346551384234095,
+      "grad_norm": 0.46634369290995936,
+      "learning_rate": 2.419909054891193e-06,
+      "loss": 0.1494,
+      "step": 2373
+    },
+    {
+      "epoch": 0.5348804461092179,
+      "grad_norm": 0.4900134894586286,
+      "learning_rate": 2.4180471346491864e-06,
+      "loss": 0.1495,
+      "step": 2374
+    },
+    {
+      "epoch": 0.5351057537950263,
+      "grad_norm": 0.46965514940495373,
+      "learning_rate": 2.4161852599125504e-06,
+      "loss": 0.1537,
+      "step": 2375
+    },
+    {
+      "epoch": 0.5353310614808348,
+      "grad_norm": 0.43801200426498405,
+      "learning_rate": 2.414323431715115e-06,
+      "loss": 0.1327,
+      "step": 2376
+    },
+    {
+      "epoch": 0.5355563691666432,
+      "grad_norm": 0.4797440345308628,
+      "learning_rate": 2.412461651090685e-06,
+      "loss": 0.1461,
+      "step": 2377
+    },
+    {
+      "epoch": 0.5357816768524516,
+      "grad_norm": 0.46511644318972317,
+      "learning_rate": 2.410599919073037e-06,
+      "loss": 0.1469,
+      "step": 2378
+    },
+    {
+      "epoch": 0.53600698453826,
+      "grad_norm": 0.47717879720300405,
+      "learning_rate": 2.408738236695922e-06,
+      "loss": 0.1514,
+      "step": 2379
+    },
+    {
+      "epoch": 0.5362322922240685,
+      "grad_norm": 0.48877338489049654,
+      "learning_rate": 2.4068766049930623e-06,
+      "loss": 0.1577,
+      "step": 2380
+    },
+    {
+      "epoch": 0.536457599909877,
+      "grad_norm": 0.49093102138390154,
+      "learning_rate": 2.4050150249981522e-06,
+      "loss": 0.1596,
+      "step": 2381
+    },
+    {
+      "epoch": 0.5366829075956854,
+      "grad_norm": 0.49852870940863864,
+      "learning_rate": 2.403153497744859e-06,
+      "loss": 0.1625,
+      "step": 2382
+    },
+    {
+      "epoch": 0.5369082152814938,
+      "grad_norm": 0.4438631646845407,
+      "learning_rate": 2.4012920242668184e-06,
+      "loss": 0.1504,
+      "step": 2383
+    },
+    {
+      "epoch": 0.5371335229673022,
+      "grad_norm": 0.490620298226887,
+      "learning_rate": 2.3994306055976374e-06,
+      "loss": 0.1589,
+      "step": 2384
+    },
+    {
+      "epoch": 0.5373588306531106,
+      "grad_norm": 0.49255329553538285,
+      "learning_rate": 2.397569242770893e-06,
+      "loss": 0.1637,
+      "step": 2385
+    },
+    {
+      "epoch": 0.537584138338919,
+      "grad_norm": 0.470517428218469,
+      "learning_rate": 2.3957079368201293e-06,
+      "loss": 0.1485,
+      "step": 2386
+    },
+    {
+      "epoch": 0.5378094460247275,
+      "grad_norm": 0.4442052068375048,
+      "learning_rate": 2.393846688778861e-06,
+      "loss": 0.1435,
+      "step": 2387
+    },
+    {
+      "epoch": 0.538034753710536,
+      "grad_norm": 0.47767306805174803,
+      "learning_rate": 2.39198549968057e-06,
+      "loss": 0.1666,
+      "step": 2388
+    },
+    {
+      "epoch": 0.5382600613963444,
+      "grad_norm": 0.4893993150625802,
+      "learning_rate": 2.390124370558705e-06,
+      "loss": 0.153,
+      "step": 2389
+    },
+    {
+      "epoch": 0.5384853690821528,
+      "grad_norm": 0.5095806079847959,
+      "learning_rate": 2.3882633024466813e-06,
+      "loss": 0.1669,
+      "step": 2390
+    },
+    {
+      "epoch": 0.5387106767679612,
+      "grad_norm": 0.461764023336289,
+      "learning_rate": 2.386402296377881e-06,
+      "loss": 0.1516,
+      "step": 2391
+    },
+    {
+      "epoch": 0.5389359844537697,
+      "grad_norm": 0.45256030151813437,
+      "learning_rate": 2.3845413533856517e-06,
+      "loss": 0.1405,
+      "step": 2392
+    },
+    {
+      "epoch": 0.5391612921395781,
+      "grad_norm": 0.48555523539900436,
+      "learning_rate": 2.3826804745033046e-06,
+      "loss": 0.1651,
+      "step": 2393
+    },
+    {
+      "epoch": 0.5393865998253865,
+      "grad_norm": 0.46075267816473214,
+      "learning_rate": 2.3808196607641176e-06,
+      "loss": 0.1522,
+      "step": 2394
+    },
+    {
+      "epoch": 0.5396119075111949,
+      "grad_norm": 0.48391583028397295,
+      "learning_rate": 2.3789589132013304e-06,
+      "loss": 0.1602,
+      "step": 2395
+    },
+    {
+      "epoch": 0.5398372151970035,
+      "grad_norm": 0.47602108556598505,
+      "learning_rate": 2.3770982328481464e-06,
+      "loss": 0.1616,
+      "step": 2396
+    },
+    {
+      "epoch": 0.5400625228828119,
+      "grad_norm": 0.4731398077638761,
+      "learning_rate": 2.3752376207377333e-06,
+      "loss": 0.1629,
+      "step": 2397
+    },
+    {
+      "epoch": 0.5402878305686203,
+      "grad_norm": 0.4678881070699935,
+      "learning_rate": 2.3733770779032185e-06,
+      "loss": 0.1578,
+      "step": 2398
+    },
+    {
+      "epoch": 0.5405131382544287,
+      "grad_norm": 0.5047146203164565,
+      "learning_rate": 2.371516605377693e-06,
+      "loss": 0.1649,
+      "step": 2399
+    },
+    {
+      "epoch": 0.5407384459402371,
+      "grad_norm": 0.46549000580913674,
+      "learning_rate": 2.3696562041942076e-06,
+      "loss": 0.161,
+      "step": 2400
+    },
+    {
+      "epoch": 0.5409637536260455,
+      "grad_norm": 0.4772724936234673,
+      "learning_rate": 2.367795875385773e-06,
+      "loss": 0.154,
+      "step": 2401
+    },
+    {
+      "epoch": 0.541189061311854,
+      "grad_norm": 0.47575796244422874,
+      "learning_rate": 2.3659356199853617e-06,
+      "loss": 0.1376,
+      "step": 2402
+    },
+    {
+      "epoch": 0.5414143689976625,
+      "grad_norm": 0.47979560555492917,
+      "learning_rate": 2.3640754390259026e-06,
+      "loss": 0.1556,
+      "step": 2403
+    },
+    {
+      "epoch": 0.5416396766834709,
+      "grad_norm": 0.47076199460274065,
+      "learning_rate": 2.362215333540287e-06,
+      "loss": 0.1571,
+      "step": 2404
+    },
+    {
+      "epoch": 0.5418649843692793,
+      "grad_norm": 0.4736130870197844,
+      "learning_rate": 2.36035530456136e-06,
+      "loss": 0.1491,
+      "step": 2405
+    },
+    {
+      "epoch": 0.5420902920550877,
+      "grad_norm": 0.4561576715451127,
+      "learning_rate": 2.3584953531219278e-06,
+      "loss": 0.1487,
+      "step": 2406
+    },
+    {
+      "epoch": 0.5423155997408962,
+      "grad_norm": 0.49891426407957157,
+      "learning_rate": 2.356635480254751e-06,
+      "loss": 0.1644,
+      "step": 2407
+    },
+    {
+      "epoch": 0.5425409074267046,
+      "grad_norm": 0.4801905005277808,
+      "learning_rate": 2.3547756869925485e-06,
+      "loss": 0.1508,
+      "step": 2408
+    },
+    {
+      "epoch": 0.542766215112513,
+      "grad_norm": 0.5022208650888927,
+      "learning_rate": 2.3529159743679936e-06,
+      "loss": 0.1625,
+      "step": 2409
+    },
+    {
+      "epoch": 0.5429915227983214,
+      "grad_norm": 0.46139824228014115,
+      "learning_rate": 2.3510563434137175e-06,
+      "loss": 0.1436,
+      "step": 2410
+    },
+    {
+      "epoch": 0.5432168304841299,
+      "grad_norm": 0.4559106153860657,
+      "learning_rate": 2.3491967951623006e-06,
+      "loss": 0.1388,
+      "step": 2411
+    },
+    {
+      "epoch": 0.5434421381699384,
+      "grad_norm": 0.4617151774727058,
+      "learning_rate": 2.347337330646282e-06,
+      "loss": 0.1518,
+      "step": 2412
+    },
+    {
+      "epoch": 0.5436674458557468,
+      "grad_norm": 0.448814934539756,
+      "learning_rate": 2.3454779508981536e-06,
+      "loss": 0.1442,
+      "step": 2413
+    },
+    {
+      "epoch": 0.5438927535415552,
+      "grad_norm": 0.5048961507681691,
+      "learning_rate": 2.3436186569503598e-06,
+      "loss": 0.1582,
+      "step": 2414
+    },
+    {
+      "epoch": 0.5441180612273636,
+      "grad_norm": 0.48880538543679386,
+      "learning_rate": 2.341759449835297e-06,
+      "loss": 0.156,
+      "step": 2415
+    },
+    {
+      "epoch": 0.544343368913172,
+      "grad_norm": 0.46704869155160356,
+      "learning_rate": 2.339900330585313e-06,
+      "loss": 0.1546,
+      "step": 2416
+    },
+    {
+      "epoch": 0.5445686765989805,
+      "grad_norm": 0.4270795493273556,
+      "learning_rate": 2.338041300232708e-06,
+      "loss": 0.1398,
+      "step": 2417
+    },
+    {
+      "epoch": 0.5447939842847889,
+      "grad_norm": 0.4543419706489556,
+      "learning_rate": 2.3361823598097316e-06,
+      "loss": 0.148,
+      "step": 2418
+    },
+    {
+      "epoch": 0.5450192919705974,
+      "grad_norm": 0.49543645533648134,
+      "learning_rate": 2.334323510348585e-06,
+      "loss": 0.1585,
+      "step": 2419
+    },
+    {
+      "epoch": 0.5452445996564058,
+      "grad_norm": 0.48534446537771797,
+      "learning_rate": 2.332464752881418e-06,
+      "loss": 0.1581,
+      "step": 2420
+    },
+    {
+      "epoch": 0.5454699073422142,
+      "grad_norm": 0.46715254152261837,
+      "learning_rate": 2.330606088440329e-06,
+      "loss": 0.1471,
+      "step": 2421
+    },
+    {
+      "epoch": 0.5456952150280227,
+      "grad_norm": 0.49929947864432916,
+      "learning_rate": 2.3287475180573653e-06,
+      "loss": 0.1536,
+      "step": 2422
+    },
+    {
+      "epoch": 0.5459205227138311,
+      "grad_norm": 0.45143500817783183,
+      "learning_rate": 2.3268890427645213e-06,
+      "loss": 0.1394,
+      "step": 2423
+    },
+    {
+      "epoch": 0.5461458303996395,
+      "grad_norm": 0.47672250458243587,
+      "learning_rate": 2.3250306635937385e-06,
+      "loss": 0.1409,
+      "step": 2424
+    },
+    {
+      "epoch": 0.5463711380854479,
+      "grad_norm": 0.48041838203924997,
+      "learning_rate": 2.323172381576907e-06,
+      "loss": 0.1516,
+      "step": 2425
+    },
+    {
+      "epoch": 0.5465964457712563,
+      "grad_norm": 0.4613411287053727,
+      "learning_rate": 2.3213141977458615e-06,
+      "loss": 0.1415,
+      "step": 2426
+    },
+    {
+      "epoch": 0.5468217534570649,
+      "grad_norm": 0.4860898207811665,
+      "learning_rate": 2.3194561131323823e-06,
+      "loss": 0.1548,
+      "step": 2427
+    },
+    {
+      "epoch": 0.5470470611428733,
+      "grad_norm": 0.5168493415498433,
+      "learning_rate": 2.3175981287681924e-06,
+      "loss": 0.1627,
+      "step": 2428
+    },
+    {
+      "epoch": 0.5472723688286817,
+      "grad_norm": 0.4950004563190919,
+      "learning_rate": 2.3157402456849632e-06,
+      "loss": 0.1559,
+      "step": 2429
+    },
+    {
+      "epoch": 0.5474976765144901,
+      "grad_norm": 0.5077308120636856,
+      "learning_rate": 2.3138824649143076e-06,
+      "loss": 0.154,
+      "step": 2430
+    },
+    {
+      "epoch": 0.5477229842002985,
+      "grad_norm": 0.4440272901386497,
+      "learning_rate": 2.312024787487782e-06,
+      "loss": 0.1498,
+      "step": 2431
+    },
+    {
+      "epoch": 0.5479482918861069,
+      "grad_norm": 0.4799839361928191,
+      "learning_rate": 2.310167214436885e-06,
+      "loss": 0.1517,
+      "step": 2432
+    },
+    {
+      "epoch": 0.5481735995719154,
+      "grad_norm": 0.49434889067719406,
+      "learning_rate": 2.3083097467930583e-06,
+      "loss": 0.1584,
+      "step": 2433
+    },
+    {
+      "epoch": 0.5483989072577238,
+      "grad_norm": 0.485517757580691,
+      "learning_rate": 2.306452385587683e-06,
+      "loss": 0.1499,
+      "step": 2434
+    },
+    {
+      "epoch": 0.5486242149435323,
+      "grad_norm": 0.4972321054541833,
+      "learning_rate": 2.304595131852085e-06,
+      "loss": 0.156,
+      "step": 2435
+    },
+    {
+      "epoch": 0.5488495226293407,
+      "grad_norm": 0.48890762164687734,
+      "learning_rate": 2.3027379866175263e-06,
+      "loss": 0.1598,
+      "step": 2436
+    },
+    {
+      "epoch": 0.5490748303151491,
+      "grad_norm": 0.4606463681481626,
+      "learning_rate": 2.300880950915211e-06,
+      "loss": 0.1497,
+      "step": 2437
+    },
+    {
+      "epoch": 0.5493001380009576,
+      "grad_norm": 0.4696087783219515,
+      "learning_rate": 2.2990240257762817e-06,
+      "loss": 0.151,
+      "step": 2438
+    },
+    {
+      "epoch": 0.549525445686766,
+      "grad_norm": 0.46354615609402217,
+      "learning_rate": 2.2971672122318196e-06,
+      "loss": 0.1296,
+      "step": 2439
+    },
+    {
+      "epoch": 0.5497507533725744,
+      "grad_norm": 0.4538811159223632,
+      "learning_rate": 2.295310511312844e-06,
+      "loss": 0.1485,
+      "step": 2440
+    },
+    {
+      "epoch": 0.5499760610583828,
+      "grad_norm": 0.5031836245171942,
+      "learning_rate": 2.293453924050312e-06,
+      "loss": 0.1542,
+      "step": 2441
+    },
+    {
+      "epoch": 0.5502013687441912,
+      "grad_norm": 0.4862351352432136,
+      "learning_rate": 2.2915974514751173e-06,
+      "loss": 0.1607,
+      "step": 2442
+    },
+    {
+      "epoch": 0.5504266764299998,
+      "grad_norm": 0.46367126664664826,
+      "learning_rate": 2.2897410946180897e-06,
+      "loss": 0.1558,
+      "step": 2443
+    },
+    {
+      "epoch": 0.5506519841158082,
+      "grad_norm": 0.4954802060102958,
+      "learning_rate": 2.287884854509995e-06,
+      "loss": 0.1476,
+      "step": 2444
+    },
+    {
+      "epoch": 0.5508772918016166,
+      "grad_norm": 0.4713809372504246,
+      "learning_rate": 2.286028732181535e-06,
+      "loss": 0.1584,
+      "step": 2445
+    },
+    {
+      "epoch": 0.551102599487425,
+      "grad_norm": 0.4838974999326031,
+      "learning_rate": 2.2841727286633444e-06,
+      "loss": 0.1486,
+      "step": 2446
+    },
+    {
+      "epoch": 0.5513279071732334,
+      "grad_norm": 0.4721516817905157,
+      "learning_rate": 2.282316844985992e-06,
+      "loss": 0.1453,
+      "step": 2447
+    },
+    {
+      "epoch": 0.5515532148590419,
+      "grad_norm": 0.4846307488266691,
+      "learning_rate": 2.280461082179982e-06,
+      "loss": 0.1511,
+      "step": 2448
+    },
+    {
+      "epoch": 0.5517785225448503,
+      "grad_norm": 0.4830182185008593,
+      "learning_rate": 2.27860544127575e-06,
+      "loss": 0.1617,
+      "step": 2449
+    },
+    {
+      "epoch": 0.5520038302306588,
+      "grad_norm": 0.4733189597116919,
+      "learning_rate": 2.2767499233036635e-06,
+      "loss": 0.1486,
+      "step": 2450
+    },
+    {
+      "epoch": 0.5522291379164672,
+      "grad_norm": 0.474466806300033,
+      "learning_rate": 2.2748945292940237e-06,
+      "loss": 0.1437,
+      "step": 2451
+    },
+    {
+      "epoch": 0.5524544456022756,
+      "grad_norm": 0.47381411512019134,
+      "learning_rate": 2.2730392602770617e-06,
+      "loss": 0.1487,
+      "step": 2452
+    },
+    {
+      "epoch": 0.552679753288084,
+      "grad_norm": 0.46505376816656113,
+      "learning_rate": 2.271184117282938e-06,
+      "loss": 0.1449,
+      "step": 2453
+    },
+    {
+      "epoch": 0.5529050609738925,
+      "grad_norm": 0.4947855734546613,
+      "learning_rate": 2.269329101341745e-06,
+      "loss": 0.1599,
+      "step": 2454
+    },
+    {
+      "epoch": 0.5531303686597009,
+      "grad_norm": 0.45848960827036234,
+      "learning_rate": 2.267474213483505e-06,
+      "loss": 0.14,
+      "step": 2455
+    },
+    {
+      "epoch": 0.5533556763455093,
+      "grad_norm": 0.47049641172903567,
+      "learning_rate": 2.265619454738166e-06,
+      "loss": 0.1506,
+      "step": 2456
+    },
+    {
+      "epoch": 0.5535809840313177,
+      "grad_norm": 0.4852049885196248,
+      "learning_rate": 2.2637648261356078e-06,
+      "loss": 0.1582,
+      "step": 2457
+    },
+    {
+      "epoch": 0.5538062917171263,
+      "grad_norm": 0.469824519560325,
+      "learning_rate": 2.2619103287056366e-06,
+      "loss": 0.1514,
+      "step": 2458
+    },
+    {
+      "epoch": 0.5540315994029347,
+      "grad_norm": 0.47447240106880856,
+      "learning_rate": 2.260055963477985e-06,
+      "loss": 0.1555,
+      "step": 2459
+    },
+    {
+      "epoch": 0.5542569070887431,
+      "grad_norm": 0.473070074615829,
+      "learning_rate": 2.2582017314823135e-06,
+      "loss": 0.1534,
+      "step": 2460
+    },
+    {
+      "epoch": 0.5544822147745515,
+      "grad_norm": 0.46559527370566156,
+      "learning_rate": 2.2563476337482073e-06,
+      "loss": 0.1543,
+      "step": 2461
+    },
+    {
+      "epoch": 0.5547075224603599,
+      "grad_norm": 0.45421710995779324,
+      "learning_rate": 2.254493671305179e-06,
+      "loss": 0.1517,
+      "step": 2462
+    },
+    {
+      "epoch": 0.5549328301461683,
+      "grad_norm": 0.4869363376822142,
+      "learning_rate": 2.2526398451826638e-06,
+      "loss": 0.1576,
+      "step": 2463
+    },
+    {
+      "epoch": 0.5551581378319768,
+      "grad_norm": 0.46141489622645343,
+      "learning_rate": 2.250786156410022e-06,
+      "loss": 0.1536,
+      "step": 2464
+    },
+    {
+      "epoch": 0.5553834455177852,
+      "grad_norm": 0.4943768008406088,
+      "learning_rate": 2.2489326060165384e-06,
+      "loss": 0.1626,
+      "step": 2465
+    },
+    {
+      "epoch": 0.5556087532035937,
+      "grad_norm": 0.4709770438969968,
+      "learning_rate": 2.24707919503142e-06,
+      "loss": 0.1553,
+      "step": 2466
+    },
+    {
+      "epoch": 0.5558340608894021,
+      "grad_norm": 0.44988307346347045,
+      "learning_rate": 2.2452259244837974e-06,
+      "loss": 0.1408,
+      "step": 2467
+    },
+    {
+      "epoch": 0.5560593685752105,
+      "grad_norm": 0.44722088845974944,
+      "learning_rate": 2.2433727954027227e-06,
+      "loss": 0.1404,
+      "step": 2468
+    },
+    {
+      "epoch": 0.556284676261019,
+      "grad_norm": 0.4842500435166934,
+      "learning_rate": 2.24151980881717e-06,
+      "loss": 0.1537,
+      "step": 2469
+    },
+    {
+      "epoch": 0.5565099839468274,
+      "grad_norm": 0.471125979789008,
+      "learning_rate": 2.239666965756032e-06,
+      "loss": 0.1434,
+      "step": 2470
+    },
+    {
+      "epoch": 0.5567352916326358,
+      "grad_norm": 0.45948464636699554,
+      "learning_rate": 2.237814267248124e-06,
+      "loss": 0.1442,
+      "step": 2471
+    },
+    {
+      "epoch": 0.5569605993184442,
+      "grad_norm": 0.48146904571655486,
+      "learning_rate": 2.2359617143221805e-06,
+      "loss": 0.1557,
+      "step": 2472
+    },
+    {
+      "epoch": 0.5571859070042526,
+      "grad_norm": 0.46022255259287337,
+      "learning_rate": 2.234109308006856e-06,
+      "loss": 0.1452,
+      "step": 2473
+    },
+    {
+      "epoch": 0.5574112146900612,
+      "grad_norm": 0.4719622400103109,
+      "learning_rate": 2.232257049330722e-06,
+      "loss": 0.1368,
+      "step": 2474
+    },
+    {
+      "epoch": 0.5576365223758696,
+      "grad_norm": 0.4698878977583087,
+      "learning_rate": 2.2304049393222686e-06,
+      "loss": 0.1494,
+      "step": 2475
+    },
+    {
+      "epoch": 0.557861830061678,
+      "grad_norm": 0.4714898445421254,
+      "learning_rate": 2.2285529790099034e-06,
+      "loss": 0.1474,
+      "step": 2476
+    },
+    {
+      "epoch": 0.5580871377474864,
+      "grad_norm": 0.47770540573100884,
+      "learning_rate": 2.2267011694219513e-06,
+      "loss": 0.1567,
+      "step": 2477
+    },
+    {
+      "epoch": 0.5583124454332948,
+      "grad_norm": 0.4909679540323867,
+      "learning_rate": 2.224849511586652e-06,
+      "loss": 0.1474,
+      "step": 2478
+    },
+    {
+      "epoch": 0.5585377531191033,
+      "grad_norm": 0.47098216682326777,
+      "learning_rate": 2.2229980065321636e-06,
+      "loss": 0.1364,
+      "step": 2479
+    },
+    {
+      "epoch": 0.5587630608049117,
+      "grad_norm": 0.4768020498354351,
+      "learning_rate": 2.221146655286558e-06,
+      "loss": 0.1599,
+      "step": 2480
+    },
+    {
+      "epoch": 0.5589883684907201,
+      "grad_norm": 0.4800137909514748,
+      "learning_rate": 2.2192954588778195e-06,
+      "loss": 0.157,
+      "step": 2481
+    },
+    {
+      "epoch": 0.5592136761765286,
+      "grad_norm": 0.4677165542238797,
+      "learning_rate": 2.21744441833385e-06,
+      "loss": 0.1532,
+      "step": 2482
+    },
+    {
+      "epoch": 0.559438983862337,
+      "grad_norm": 0.487362845686037,
+      "learning_rate": 2.2155935346824634e-06,
+      "loss": 0.1625,
+      "step": 2483
+    },
+    {
+      "epoch": 0.5596642915481455,
+      "grad_norm": 0.4736455775888925,
+      "learning_rate": 2.2137428089513857e-06,
+      "loss": 0.1476,
+      "step": 2484
+    },
+    {
+      "epoch": 0.5598895992339539,
+      "grad_norm": 0.5085683208858834,
+      "learning_rate": 2.2118922421682563e-06,
+      "loss": 0.1628,
+      "step": 2485
+    },
+    {
+      "epoch": 0.5601149069197623,
+      "grad_norm": 0.4795859420103309,
+      "learning_rate": 2.2100418353606262e-06,
+      "loss": 0.1561,
+      "step": 2486
+    },
+    {
+      "epoch": 0.5603402146055707,
+      "grad_norm": 0.4727352372361001,
+      "learning_rate": 2.208191589555956e-06,
+      "loss": 0.1433,
+      "step": 2487
+    },
+    {
+      "epoch": 0.5605655222913791,
+      "grad_norm": 0.4978582284269819,
+      "learning_rate": 2.20634150578162e-06,
+      "loss": 0.1529,
+      "step": 2488
+    },
+    {
+      "epoch": 0.5607908299771875,
+      "grad_norm": 0.4799519585661368,
+      "learning_rate": 2.2044915850649e-06,
+      "loss": 0.1495,
+      "step": 2489
+    },
+    {
+      "epoch": 0.5610161376629961,
+      "grad_norm": 0.4882231226390542,
+      "learning_rate": 2.202641828432988e-06,
+      "loss": 0.1538,
+      "step": 2490
+    },
+    {
+      "epoch": 0.5612414453488045,
+      "grad_norm": 0.49437833162263073,
+      "learning_rate": 2.2007922369129854e-06,
+      "loss": 0.1629,
+      "step": 2491
+    },
+    {
+      "epoch": 0.5614667530346129,
+      "grad_norm": 0.48283606695640197,
+      "learning_rate": 2.1989428115319005e-06,
+      "loss": 0.1559,
+      "step": 2492
+    },
+    {
+      "epoch": 0.5616920607204213,
+      "grad_norm": 0.4824869865618337,
+      "learning_rate": 2.1970935533166505e-06,
+      "loss": 0.1551,
+      "step": 2493
+    },
+    {
+      "epoch": 0.5619173684062297,
+      "grad_norm": 0.47903485883095387,
+      "learning_rate": 2.19524446329406e-06,
+      "loss": 0.1464,
+      "step": 2494
+    },
+    {
+      "epoch": 0.5621426760920382,
+      "grad_norm": 0.47432503827693573,
+      "learning_rate": 2.1933955424908596e-06,
+      "loss": 0.164,
+      "step": 2495
+    },
+    {
+      "epoch": 0.5623679837778466,
+      "grad_norm": 0.44292504233636465,
+      "learning_rate": 2.1915467919336862e-06,
+      "loss": 0.1338,
+      "step": 2496
+    },
+    {
+      "epoch": 0.562593291463655,
+      "grad_norm": 0.4519280759224281,
+      "learning_rate": 2.1896982126490825e-06,
+      "loss": 0.1498,
+      "step": 2497
+    },
+    {
+      "epoch": 0.5628185991494635,
+      "grad_norm": 0.49827045995047065,
+      "learning_rate": 2.1878498056634946e-06,
+      "loss": 0.1618,
+      "step": 2498
+    },
+    {
+      "epoch": 0.563043906835272,
+      "grad_norm": 0.4730814583558337,
+      "learning_rate": 2.1860015720032747e-06,
+      "loss": 0.1539,
+      "step": 2499
+    },
+    {
+      "epoch": 0.5632692145210804,
+      "grad_norm": 0.4851839051181484,
+      "learning_rate": 2.1841535126946777e-06,
+      "loss": 0.1425,
+      "step": 2500
+    },
+    {
+      "epoch": 0.5632692145210804,
+      "eval_loss": 0.15211372077465057,
+      "eval_runtime": 56.972,
+      "eval_samples_per_second": 50.376,
+      "eval_steps_per_second": 6.301,
+      "step": 2500
+    },
+    {
+      "epoch": 0.5634945222068888,
+      "grad_norm": 0.5224387250101071,
+      "learning_rate": 2.1823056287638623e-06,
+      "loss": 0.1621,
+      "step": 2501
+    },
+    {
+      "epoch": 0.5637198298926972,
+      "grad_norm": 0.4877748448316099,
+      "learning_rate": 2.180457921236889e-06,
+      "loss": 0.1497,
+      "step": 2502
+    },
+    {
+      "epoch": 0.5639451375785056,
+      "grad_norm": 0.4737044072894314,
+      "learning_rate": 2.1786103911397218e-06,
+      "loss": 0.1477,
+      "step": 2503
+    },
+    {
+      "epoch": 0.564170445264314,
+      "grad_norm": 0.44795906173065436,
+      "learning_rate": 2.176763039498225e-06,
+      "loss": 0.1452,
+      "step": 2504
+    },
+    {
+      "epoch": 0.5643957529501226,
+      "grad_norm": 0.48310843295660033,
+      "learning_rate": 2.174915867338164e-06,
+      "loss": 0.1458,
+      "step": 2505
+    },
+    {
+      "epoch": 0.564621060635931,
+      "grad_norm": 0.47537092410064236,
+      "learning_rate": 2.1730688756852046e-06,
+      "loss": 0.1497,
+      "step": 2506
+    },
+    {
+      "epoch": 0.5648463683217394,
+      "grad_norm": 0.4872498445237118,
+      "learning_rate": 2.171222065564913e-06,
+      "loss": 0.1656,
+      "step": 2507
+    },
+    {
+      "epoch": 0.5650716760075478,
+      "grad_norm": 0.4905041228342476,
+      "learning_rate": 2.1693754380027533e-06,
+      "loss": 0.1563,
+      "step": 2508
+    },
+    {
+      "epoch": 0.5652969836933562,
+      "grad_norm": 0.47516862562513035,
+      "learning_rate": 2.1675289940240883e-06,
+      "loss": 0.1424,
+      "step": 2509
+    },
+    {
+      "epoch": 0.5655222913791647,
+      "grad_norm": 0.42823728751408857,
+      "learning_rate": 2.165682734654181e-06,
+      "loss": 0.1282,
+      "step": 2510
+    },
+    {
+      "epoch": 0.5657475990649731,
+      "grad_norm": 0.5124273639117506,
+      "learning_rate": 2.16383666091819e-06,
+      "loss": 0.1566,
+      "step": 2511
+    },
+    {
+      "epoch": 0.5659729067507815,
+      "grad_norm": 0.4662028274966955,
+      "learning_rate": 2.161990773841171e-06,
+      "loss": 0.1547,
+      "step": 2512
+    },
+    {
+      "epoch": 0.56619821443659,
+      "grad_norm": 0.47300391524406277,
+      "learning_rate": 2.1601450744480763e-06,
+      "loss": 0.1499,
+      "step": 2513
+    },
+    {
+      "epoch": 0.5664235221223984,
+      "grad_norm": 0.5055930403538434,
+      "learning_rate": 2.1582995637637543e-06,
+      "loss": 0.1602,
+      "step": 2514
+    },
+    {
+      "epoch": 0.5666488298082069,
+      "grad_norm": 0.46885317458281295,
+      "learning_rate": 2.156454242812948e-06,
+      "loss": 0.1429,
+      "step": 2515
+    },
+    {
+      "epoch": 0.5668741374940153,
+      "grad_norm": 0.4912017132958433,
+      "learning_rate": 2.1546091126202955e-06,
+      "loss": 0.1636,
+      "step": 2516
+    },
+    {
+      "epoch": 0.5670994451798237,
+      "grad_norm": 0.4595886886992803,
+      "learning_rate": 2.1527641742103282e-06,
+      "loss": 0.1405,
+      "step": 2517
+    },
+    {
+      "epoch": 0.5673247528656321,
+      "grad_norm": 0.4988690329908019,
+      "learning_rate": 2.150919428607472e-06,
+      "loss": 0.1501,
+      "step": 2518
+    },
+    {
+      "epoch": 0.5675500605514405,
+      "grad_norm": 0.4762729304575605,
+      "learning_rate": 2.149074876836045e-06,
+      "loss": 0.1564,
+      "step": 2519
+    },
+    {
+      "epoch": 0.567775368237249,
+      "grad_norm": 0.47204168342815955,
+      "learning_rate": 2.147230519920259e-06,
+      "loss": 0.1641,
+      "step": 2520
+    },
+    {
+      "epoch": 0.5680006759230575,
+      "grad_norm": 0.46207272776568986,
+      "learning_rate": 2.1453863588842165e-06,
+      "loss": 0.1418,
+      "step": 2521
+    },
+    {
+      "epoch": 0.5682259836088659,
+      "grad_norm": 0.4668825504120351,
+      "learning_rate": 2.143542394751911e-06,
+      "loss": 0.1538,
+      "step": 2522
+    },
+    {
+      "epoch": 0.5684512912946743,
+      "grad_norm": 0.4837268539344098,
+      "learning_rate": 2.1416986285472268e-06,
+      "loss": 0.1541,
+      "step": 2523
+    },
+    {
+      "epoch": 0.5686765989804827,
+      "grad_norm": 0.4624010885176109,
+      "learning_rate": 2.139855061293939e-06,
+      "loss": 0.1537,
+      "step": 2524
+    },
+    {
+      "epoch": 0.5689019066662911,
+      "grad_norm": 0.4710709274529512,
+      "learning_rate": 2.1380116940157107e-06,
+      "loss": 0.155,
+      "step": 2525
+    },
+    {
+      "epoch": 0.5691272143520996,
+      "grad_norm": 0.5156089057170854,
+      "learning_rate": 2.1361685277360973e-06,
+      "loss": 0.1608,
+      "step": 2526
+    },
+    {
+      "epoch": 0.569352522037908,
+      "grad_norm": 0.48493200248859675,
+      "learning_rate": 2.1343255634785386e-06,
+      "loss": 0.1503,
+      "step": 2527
+    },
+    {
+      "epoch": 0.5695778297237164,
+      "grad_norm": 0.4333440828139637,
+      "learning_rate": 2.132482802266364e-06,
+      "loss": 0.122,
+      "step": 2528
+    },
+    {
+      "epoch": 0.5698031374095249,
+      "grad_norm": 0.4607815473590659,
+      "learning_rate": 2.1306402451227907e-06,
+      "loss": 0.1503,
+      "step": 2529
+    },
+    {
+      "epoch": 0.5700284450953333,
+      "grad_norm": 0.46389727276493975,
+      "learning_rate": 2.128797893070922e-06,
+      "loss": 0.1513,
+      "step": 2530
+    },
+    {
+      "epoch": 0.5702537527811418,
+      "grad_norm": 0.5132583394666452,
+      "learning_rate": 2.1269557471337467e-06,
+      "loss": 0.1515,
+      "step": 2531
+    },
+    {
+      "epoch": 0.5704790604669502,
+      "grad_norm": 0.5026188029568095,
+      "learning_rate": 2.1251138083341404e-06,
+      "loss": 0.1528,
+      "step": 2532
+    },
+    {
+      "epoch": 0.5707043681527586,
+      "grad_norm": 0.481269253549719,
+      "learning_rate": 2.123272077694864e-06,
+      "loss": 0.1453,
+      "step": 2533
+    },
+    {
+      "epoch": 0.570929675838567,
+      "grad_norm": 0.46800101724365095,
+      "learning_rate": 2.1214305562385592e-06,
+      "loss": 0.1548,
+      "step": 2534
+    },
+    {
+      "epoch": 0.5711549835243754,
+      "grad_norm": 0.4658194574312725,
+      "learning_rate": 2.1195892449877556e-06,
+      "loss": 0.1455,
+      "step": 2535
+    },
+    {
+      "epoch": 0.5713802912101839,
+      "grad_norm": 0.4773496273400579,
+      "learning_rate": 2.117748144964865e-06,
+      "loss": 0.1571,
+      "step": 2536
+    },
+    {
+      "epoch": 0.5716055988959924,
+      "grad_norm": 0.4741653276708234,
+      "learning_rate": 2.115907257192182e-06,
+      "loss": 0.1482,
+      "step": 2537
+    },
+    {
+      "epoch": 0.5718309065818008,
+      "grad_norm": 0.4492929001791757,
+      "learning_rate": 2.1140665826918823e-06,
+      "loss": 0.138,
+      "step": 2538
+    },
+    {
+      "epoch": 0.5720562142676092,
+      "grad_norm": 0.5064130427481476,
+      "learning_rate": 2.1122261224860237e-06,
+      "loss": 0.1645,
+      "step": 2539
+    },
+    {
+      "epoch": 0.5722815219534176,
+      "grad_norm": 0.46568785233626575,
+      "learning_rate": 2.1103858775965455e-06,
+      "loss": 0.1564,
+      "step": 2540
+    },
+    {
+      "epoch": 0.5725068296392261,
+      "grad_norm": 0.4824316713930859,
+      "learning_rate": 2.1085458490452663e-06,
+      "loss": 0.1539,
+      "step": 2541
+    },
+    {
+      "epoch": 0.5727321373250345,
+      "grad_norm": 0.48853412799556045,
+      "learning_rate": 2.106706037853887e-06,
+      "loss": 0.1441,
+      "step": 2542
+    },
+    {
+      "epoch": 0.5729574450108429,
+      "grad_norm": 0.46998661739099296,
+      "learning_rate": 2.1048664450439853e-06,
+      "loss": 0.1403,
+      "step": 2543
+    },
+    {
+      "epoch": 0.5731827526966513,
+      "grad_norm": 0.4822964462853322,
+      "learning_rate": 2.103027071637018e-06,
+      "loss": 0.149,
+      "step": 2544
+    },
+    {
+      "epoch": 0.5734080603824598,
+      "grad_norm": 0.4713654126005489,
+      "learning_rate": 2.101187918654321e-06,
+      "loss": 0.143,
+      "step": 2545
+    },
+    {
+      "epoch": 0.5736333680682683,
+      "grad_norm": 0.5069590134179427,
+      "learning_rate": 2.099348987117108e-06,
+      "loss": 0.1608,
+      "step": 2546
+    },
+    {
+      "epoch": 0.5738586757540767,
+      "grad_norm": 0.48913938668701046,
+      "learning_rate": 2.0975102780464674e-06,
+      "loss": 0.1542,
+      "step": 2547
+    },
+    {
+      "epoch": 0.5740839834398851,
+      "grad_norm": 0.4917953755983581,
+      "learning_rate": 2.095671792463368e-06,
+      "loss": 0.1588,
+      "step": 2548
+    },
+    {
+      "epoch": 0.5743092911256935,
+      "grad_norm": 0.47742831862757695,
+      "learning_rate": 2.0938335313886513e-06,
+      "loss": 0.1547,
+      "step": 2549
+    },
+    {
+      "epoch": 0.5745345988115019,
+      "grad_norm": 0.4400050925635982,
+      "learning_rate": 2.0919954958430357e-06,
+      "loss": 0.1359,
+      "step": 2550
+    },
+    {
+      "epoch": 0.5747599064973103,
+      "grad_norm": 0.5085400055192658,
+      "learning_rate": 2.0901576868471125e-06,
+      "loss": 0.1596,
+      "step": 2551
+    },
+    {
+      "epoch": 0.5749852141831189,
+      "grad_norm": 0.44727232653419413,
+      "learning_rate": 2.0883201054213493e-06,
+      "loss": 0.1451,
+      "step": 2552
+    },
+    {
+      "epoch": 0.5752105218689273,
+      "grad_norm": 0.46059727795786004,
+      "learning_rate": 2.086482752586087e-06,
+      "loss": 0.1487,
+      "step": 2553
+    },
+    {
+      "epoch": 0.5754358295547357,
+      "grad_norm": 0.46831069725509195,
+      "learning_rate": 2.0846456293615384e-06,
+      "loss": 0.1468,
+      "step": 2554
+    },
+    {
+      "epoch": 0.5756611372405441,
+      "grad_norm": 0.46378208101596174,
+      "learning_rate": 2.0828087367677906e-06,
+      "loss": 0.1518,
+      "step": 2555
+    },
+    {
+      "epoch": 0.5758864449263525,
+      "grad_norm": 0.4756688058583516,
+      "learning_rate": 2.0809720758247997e-06,
+      "loss": 0.1604,
+      "step": 2556
+    },
+    {
+      "epoch": 0.576111752612161,
+      "grad_norm": 0.5154499252124413,
+      "learning_rate": 2.0791356475523967e-06,
+      "loss": 0.1649,
+      "step": 2557
+    },
+    {
+      "epoch": 0.5763370602979694,
+      "grad_norm": 0.49943906859891457,
+      "learning_rate": 2.077299452970282e-06,
+      "loss": 0.1584,
+      "step": 2558
+    },
+    {
+      "epoch": 0.5765623679837778,
+      "grad_norm": 0.47353120279436367,
+      "learning_rate": 2.0754634930980245e-06,
+      "loss": 0.1526,
+      "step": 2559
+    },
+    {
+      "epoch": 0.5767876756695863,
+      "grad_norm": 0.4764005539219673,
+      "learning_rate": 2.0736277689550655e-06,
+      "loss": 0.1521,
+      "step": 2560
+    },
+    {
+      "epoch": 0.5770129833553947,
+      "grad_norm": 0.46881451194766166,
+      "learning_rate": 2.0717922815607134e-06,
+      "loss": 0.1569,
+      "step": 2561
+    },
+    {
+      "epoch": 0.5772382910412032,
+      "grad_norm": 0.4827888344579709,
+      "learning_rate": 2.069957031934147e-06,
+      "loss": 0.1478,
+      "step": 2562
+    },
+    {
+      "epoch": 0.5774635987270116,
+      "grad_norm": 0.5023148970873527,
+      "learning_rate": 2.0681220210944106e-06,
+      "loss": 0.1609,
+      "step": 2563
+    },
+    {
+      "epoch": 0.57768890641282,
+      "grad_norm": 0.49917019654858047,
+      "learning_rate": 2.066287250060418e-06,
+      "loss": 0.1616,
+      "step": 2564
+    },
+    {
+      "epoch": 0.5779142140986284,
+      "grad_norm": 0.4968052640321321,
+      "learning_rate": 2.06445271985095e-06,
+      "loss": 0.1575,
+      "step": 2565
+    },
+    {
+      "epoch": 0.5781395217844368,
+      "grad_norm": 0.4487199967977616,
+      "learning_rate": 2.062618431484652e-06,
+      "loss": 0.1438,
+      "step": 2566
+    },
+    {
+      "epoch": 0.5783648294702453,
+      "grad_norm": 0.4846379596071071,
+      "learning_rate": 2.060784385980036e-06,
+      "loss": 0.1493,
+      "step": 2567
+    },
+    {
+      "epoch": 0.5785901371560538,
+      "grad_norm": 0.4792229250058271,
+      "learning_rate": 2.05895058435548e-06,
+      "loss": 0.138,
+      "step": 2568
+    },
+    {
+      "epoch": 0.5788154448418622,
+      "grad_norm": 0.4901580327188909,
+      "learning_rate": 2.0571170276292233e-06,
+      "loss": 0.1469,
+      "step": 2569
+    },
+    {
+      "epoch": 0.5790407525276706,
+      "grad_norm": 0.48749294511610947,
+      "learning_rate": 2.0552837168193738e-06,
+      "loss": 0.1497,
+      "step": 2570
+    },
+    {
+      "epoch": 0.579266060213479,
+      "grad_norm": 0.4548388896156358,
+      "learning_rate": 2.0534506529439e-06,
+      "loss": 0.1419,
+      "step": 2571
+    },
+    {
+      "epoch": 0.5794913678992875,
+      "grad_norm": 0.5154923361811012,
+      "learning_rate": 2.051617837020633e-06,
+      "loss": 0.1596,
+      "step": 2572
+    },
+    {
+      "epoch": 0.5797166755850959,
+      "grad_norm": 0.4729595000797912,
+      "learning_rate": 2.0497852700672692e-06,
+      "loss": 0.14,
+      "step": 2573
+    },
+    {
+      "epoch": 0.5799419832709043,
+      "grad_norm": 0.4959258355172917,
+      "learning_rate": 2.047952953101363e-06,
+      "loss": 0.1604,
+      "step": 2574
+    },
+    {
+      "epoch": 0.5801672909567127,
+      "grad_norm": 0.45432381151595685,
+      "learning_rate": 2.0461208871403333e-06,
+      "loss": 0.1414,
+      "step": 2575
+    },
+    {
+      "epoch": 0.5803925986425212,
+      "grad_norm": 0.47096658021491633,
+      "learning_rate": 2.0442890732014563e-06,
+      "loss": 0.1454,
+      "step": 2576
+    },
+    {
+      "epoch": 0.5806179063283297,
+      "grad_norm": 0.4525308560866689,
+      "learning_rate": 2.042457512301871e-06,
+      "loss": 0.14,
+      "step": 2577
+    },
+    {
+      "epoch": 0.5808432140141381,
+      "grad_norm": 0.5020390621894459,
+      "learning_rate": 2.040626205458574e-06,
+      "loss": 0.1537,
+      "step": 2578
+    },
+    {
+      "epoch": 0.5810685216999465,
+      "grad_norm": 0.4954963757896994,
+      "learning_rate": 2.038795153688423e-06,
+      "loss": 0.1578,
+      "step": 2579
+    },
+    {
+      "epoch": 0.5812938293857549,
+      "grad_norm": 0.4928423187106702,
+      "learning_rate": 2.0369643580081326e-06,
+      "loss": 0.1594,
+      "step": 2580
+    },
+    {
+      "epoch": 0.5815191370715633,
+      "grad_norm": 0.4882262305198443,
+      "learning_rate": 2.0351338194342744e-06,
+      "loss": 0.1478,
+      "step": 2581
+    },
+    {
+      "epoch": 0.5817444447573717,
+      "grad_norm": 0.4604675050885477,
+      "learning_rate": 2.0333035389832795e-06,
+      "loss": 0.1417,
+      "step": 2582
+    },
+    {
+      "epoch": 0.5819697524431802,
+      "grad_norm": 0.5080053698936663,
+      "learning_rate": 2.0314735176714336e-06,
+      "loss": 0.1575,
+      "step": 2583
+    },
+    {
+      "epoch": 0.5821950601289887,
+      "grad_norm": 0.5019888888860421,
+      "learning_rate": 2.0296437565148786e-06,
+      "loss": 0.161,
+      "step": 2584
+    },
+    {
+      "epoch": 0.5824203678147971,
+      "grad_norm": 0.49531296644130557,
+      "learning_rate": 2.0278142565296153e-06,
+      "loss": 0.1429,
+      "step": 2585
+    },
+    {
+      "epoch": 0.5826456755006055,
+      "grad_norm": 0.4866792190717066,
+      "learning_rate": 2.025985018731494e-06,
+      "loss": 0.1581,
+      "step": 2586
+    },
+    {
+      "epoch": 0.582870983186414,
+      "grad_norm": 0.4894237643330584,
+      "learning_rate": 2.0241560441362235e-06,
+      "loss": 0.1524,
+      "step": 2587
+    },
+    {
+      "epoch": 0.5830962908722224,
+      "grad_norm": 0.4945925080814945,
+      "learning_rate": 2.0223273337593647e-06,
+      "loss": 0.1572,
+      "step": 2588
+    },
+    {
+      "epoch": 0.5833215985580308,
+      "grad_norm": 0.4973169711801826,
+      "learning_rate": 2.020498888616333e-06,
+      "loss": 0.1583,
+      "step": 2589
+    },
+    {
+      "epoch": 0.5835469062438392,
+      "grad_norm": 0.5017905379704919,
+      "learning_rate": 2.0186707097223952e-06,
+      "loss": 0.1509,
+      "step": 2590
+    },
+    {
+      "epoch": 0.5837722139296476,
+      "grad_norm": 0.48194812401972437,
+      "learning_rate": 2.016842798092672e-06,
+      "loss": 0.1458,
+      "step": 2591
+    },
+    {
+      "epoch": 0.5839975216154561,
+      "grad_norm": 0.4790546754896114,
+      "learning_rate": 2.0150151547421333e-06,
+      "loss": 0.1524,
+      "step": 2592
+    },
+    {
+      "epoch": 0.5842228293012646,
+      "grad_norm": 0.49411625346523946,
+      "learning_rate": 2.013187780685602e-06,
+      "loss": 0.1494,
+      "step": 2593
+    },
+    {
+      "epoch": 0.584448136987073,
+      "grad_norm": 0.44879665522584505,
+      "learning_rate": 2.0113606769377497e-06,
+      "loss": 0.1469,
+      "step": 2594
+    },
+    {
+      "epoch": 0.5846734446728814,
+      "grad_norm": 0.48272208567022384,
+      "learning_rate": 2.009533844513101e-06,
+      "loss": 0.1552,
+      "step": 2595
+    },
+    {
+      "epoch": 0.5848987523586898,
+      "grad_norm": 0.4599871152837736,
+      "learning_rate": 2.0077072844260267e-06,
+      "loss": 0.1406,
+      "step": 2596
+    },
+    {
+      "epoch": 0.5851240600444982,
+      "grad_norm": 0.5004012967022493,
+      "learning_rate": 2.0058809976907475e-06,
+      "loss": 0.1498,
+      "step": 2597
+    },
+    {
+      "epoch": 0.5853493677303067,
+      "grad_norm": 0.46983282686813543,
+      "learning_rate": 2.0040549853213326e-06,
+      "loss": 0.1571,
+      "step": 2598
+    },
+    {
+      "epoch": 0.5855746754161152,
+      "grad_norm": 0.46983737953382637,
+      "learning_rate": 2.0022292483316984e-06,
+      "loss": 0.1503,
+      "step": 2599
+    },
+    {
+      "epoch": 0.5857999831019236,
+      "grad_norm": 0.4820623624485316,
+      "learning_rate": 2.0004037877356085e-06,
+      "loss": 0.1516,
+      "step": 2600
+    },
+    {
+      "epoch": 0.586025290787732,
+      "grad_norm": 0.4855950313891801,
+      "learning_rate": 1.998578604546674e-06,
+      "loss": 0.1506,
+      "step": 2601
+    },
+    {
+      "epoch": 0.5862505984735404,
+      "grad_norm": 0.4875750250191104,
+      "learning_rate": 1.9967536997783495e-06,
+      "loss": 0.1578,
+      "step": 2602
+    },
+    {
+      "epoch": 0.5864759061593489,
+      "grad_norm": 0.4553425030090635,
+      "learning_rate": 1.9949290744439392e-06,
+      "loss": 0.1456,
+      "step": 2603
+    },
+    {
+      "epoch": 0.5867012138451573,
+      "grad_norm": 0.4804593658936272,
+      "learning_rate": 1.9931047295565863e-06,
+      "loss": 0.1531,
+      "step": 2604
+    },
+    {
+      "epoch": 0.5869265215309657,
+      "grad_norm": 0.45953845789303216,
+      "learning_rate": 1.9912806661292838e-06,
+      "loss": 0.1428,
+      "step": 2605
+    },
+    {
+      "epoch": 0.5871518292167741,
+      "grad_norm": 0.49028482450191974,
+      "learning_rate": 1.989456885174865e-06,
+      "loss": 0.1575,
+      "step": 2606
+    },
+    {
+      "epoch": 0.5873771369025826,
+      "grad_norm": 0.4458092542065396,
+      "learning_rate": 1.987633387706008e-06,
+      "loss": 0.1346,
+      "step": 2607
+    },
+    {
+      "epoch": 0.5876024445883911,
+      "grad_norm": 0.4609231008990843,
+      "learning_rate": 1.9858101747352326e-06,
+      "loss": 0.1426,
+      "step": 2608
+    },
+    {
+      "epoch": 0.5878277522741995,
+      "grad_norm": 0.49280751358160235,
+      "learning_rate": 1.9839872472749016e-06,
+      "loss": 0.1497,
+      "step": 2609
+    },
+    {
+      "epoch": 0.5880530599600079,
+      "grad_norm": 0.4375963099756855,
+      "learning_rate": 1.9821646063372174e-06,
+      "loss": 0.1349,
+      "step": 2610
+    },
+    {
+      "epoch": 0.5882783676458163,
+      "grad_norm": 0.4595108845111774,
+      "learning_rate": 1.9803422529342264e-06,
+      "loss": 0.1341,
+      "step": 2611
+    },
+    {
+      "epoch": 0.5885036753316247,
+      "grad_norm": 0.4526724622007013,
+      "learning_rate": 1.978520188077813e-06,
+      "loss": 0.1384,
+      "step": 2612
+    },
+    {
+      "epoch": 0.5887289830174331,
+      "grad_norm": 0.5053804717077324,
+      "learning_rate": 1.976698412779701e-06,
+      "loss": 0.1624,
+      "step": 2613
+    },
+    {
+      "epoch": 0.5889542907032416,
+      "grad_norm": 0.45473150017860553,
+      "learning_rate": 1.9748769280514544e-06,
+      "loss": 0.1468,
+      "step": 2614
+    },
+    {
+      "epoch": 0.5891795983890501,
+      "grad_norm": 0.4774154720585141,
+      "learning_rate": 1.973055734904477e-06,
+      "loss": 0.1521,
+      "step": 2615
+    },
+    {
+      "epoch": 0.5894049060748585,
+      "grad_norm": 0.4993969480390666,
+      "learning_rate": 1.971234834350008e-06,
+      "loss": 0.1493,
+      "step": 2616
+    },
+    {
+      "epoch": 0.5896302137606669,
+      "grad_norm": 0.4814371640595727,
+      "learning_rate": 1.9694142273991264e-06,
+      "loss": 0.15,
+      "step": 2617
+    },
+    {
+      "epoch": 0.5898555214464754,
+      "grad_norm": 0.4811078134010406,
+      "learning_rate": 1.967593915062748e-06,
+      "loss": 0.1549,
+      "step": 2618
+    },
+    {
+      "epoch": 0.5900808291322838,
+      "grad_norm": 0.49120183783803484,
+      "learning_rate": 1.9657738983516227e-06,
+      "loss": 0.1528,
+      "step": 2619
+    },
+    {
+      "epoch": 0.5903061368180922,
+      "grad_norm": 0.5010617162358244,
+      "learning_rate": 1.96395417827634e-06,
+      "loss": 0.1643,
+      "step": 2620
+    },
+    {
+      "epoch": 0.5905314445039006,
+      "grad_norm": 0.4265067214265054,
+      "learning_rate": 1.9621347558473216e-06,
+      "loss": 0.1257,
+      "step": 2621
+    },
+    {
+      "epoch": 0.590756752189709,
+      "grad_norm": 0.48263433337715467,
+      "learning_rate": 1.960315632074824e-06,
+      "loss": 0.1532,
+      "step": 2622
+    },
+    {
+      "epoch": 0.5909820598755176,
+      "grad_norm": 0.4806343374996929,
+      "learning_rate": 1.95849680796894e-06,
+      "loss": 0.1459,
+      "step": 2623
+    },
+    {
+      "epoch": 0.591207367561326,
+      "grad_norm": 0.484164716047899,
+      "learning_rate": 1.9566782845395945e-06,
+      "loss": 0.1572,
+      "step": 2624
+    },
+    {
+      "epoch": 0.5914326752471344,
+      "grad_norm": 0.4839391840783974,
+      "learning_rate": 1.9548600627965454e-06,
+      "loss": 0.1607,
+      "step": 2625
+    },
+    {
+      "epoch": 0.5916579829329428,
+      "grad_norm": 0.49965378539140476,
+      "learning_rate": 1.9530421437493843e-06,
+      "loss": 0.1531,
+      "step": 2626
+    },
+    {
+      "epoch": 0.5918832906187512,
+      "grad_norm": 0.4769090414318042,
+      "learning_rate": 1.951224528407534e-06,
+      "loss": 0.1454,
+      "step": 2627
+    },
+    {
+      "epoch": 0.5921085983045596,
+      "grad_norm": 0.47164774135841486,
+      "learning_rate": 1.949407217780247e-06,
+      "loss": 0.1517,
+      "step": 2628
+    },
+    {
+      "epoch": 0.5923339059903681,
+      "grad_norm": 0.4840071951346353,
+      "learning_rate": 1.94759021287661e-06,
+      "loss": 0.163,
+      "step": 2629
+    },
+    {
+      "epoch": 0.5925592136761765,
+      "grad_norm": 0.4964969178311975,
+      "learning_rate": 1.945773514705537e-06,
+      "loss": 0.1517,
+      "step": 2630
+    },
+    {
+      "epoch": 0.592784521361985,
+      "grad_norm": 0.49816050788736993,
+      "learning_rate": 1.943957124275773e-06,
+      "loss": 0.1557,
+      "step": 2631
+    },
+    {
+      "epoch": 0.5930098290477934,
+      "grad_norm": 0.4413938462386378,
+      "learning_rate": 1.9421410425958915e-06,
+      "loss": 0.1296,
+      "step": 2632
+    },
+    {
+      "epoch": 0.5932351367336018,
+      "grad_norm": 0.5117692482649495,
+      "learning_rate": 1.9403252706742957e-06,
+      "loss": 0.162,
+      "step": 2633
+    },
+    {
+      "epoch": 0.5934604444194103,
+      "grad_norm": 0.4582300780902382,
+      "learning_rate": 1.938509809519216e-06,
+      "loss": 0.1419,
+      "step": 2634
+    },
+    {
+      "epoch": 0.5936857521052187,
+      "grad_norm": 0.4811552477231402,
+      "learning_rate": 1.9366946601387103e-06,
+      "loss": 0.1483,
+      "step": 2635
+    },
+    {
+      "epoch": 0.5939110597910271,
+      "grad_norm": 0.513375477344272,
+      "learning_rate": 1.934879823540663e-06,
+      "loss": 0.1699,
+      "step": 2636
+    },
+    {
+      "epoch": 0.5941363674768355,
+      "grad_norm": 0.5216389103084176,
+      "learning_rate": 1.9330653007327852e-06,
+      "loss": 0.162,
+      "step": 2637
+    },
+    {
+      "epoch": 0.5943616751626439,
+      "grad_norm": 0.4702534040240756,
+      "learning_rate": 1.931251092722615e-06,
+      "loss": 0.1518,
+      "step": 2638
+    },
+    {
+      "epoch": 0.5945869828484525,
+      "grad_norm": 0.4873357715988879,
+      "learning_rate": 1.9294372005175125e-06,
+      "loss": 0.1514,
+      "step": 2639
+    },
+    {
+      "epoch": 0.5948122905342609,
+      "grad_norm": 0.5050029824413044,
+      "learning_rate": 1.9276236251246655e-06,
+      "loss": 0.1496,
+      "step": 2640
+    },
+    {
+      "epoch": 0.5950375982200693,
+      "grad_norm": 0.45566476221772784,
+      "learning_rate": 1.9258103675510846e-06,
+      "loss": 0.1444,
+      "step": 2641
+    },
+    {
+      "epoch": 0.5952629059058777,
+      "grad_norm": 0.5160389746563608,
+      "learning_rate": 1.9239974288036044e-06,
+      "loss": 0.1655,
+      "step": 2642
+    },
+    {
+      "epoch": 0.5954882135916861,
+      "grad_norm": 0.4321268048853623,
+      "learning_rate": 1.9221848098888817e-06,
+      "loss": 0.1282,
+      "step": 2643
+    },
+    {
+      "epoch": 0.5957135212774946,
+      "grad_norm": 0.4705484540912023,
+      "learning_rate": 1.920372511813397e-06,
+      "loss": 0.1418,
+      "step": 2644
+    },
+    {
+      "epoch": 0.595938828963303,
+      "grad_norm": 0.4727463487631332,
+      "learning_rate": 1.9185605355834518e-06,
+      "loss": 0.1495,
+      "step": 2645
+    },
+    {
+      "epoch": 0.5961641366491115,
+      "grad_norm": 0.47996660778579237,
+      "learning_rate": 1.916748882205168e-06,
+      "loss": 0.1393,
+      "step": 2646
+    },
+    {
+      "epoch": 0.5963894443349199,
+      "grad_norm": 0.47144733107356446,
+      "learning_rate": 1.91493755268449e-06,
+      "loss": 0.1427,
+      "step": 2647
+    },
+    {
+      "epoch": 0.5966147520207283,
+      "grad_norm": 0.4486714323008961,
+      "learning_rate": 1.913126548027181e-06,
+      "loss": 0.1444,
+      "step": 2648
+    },
+    {
+      "epoch": 0.5968400597065368,
+      "grad_norm": 0.46483001125109336,
+      "learning_rate": 1.9113158692388253e-06,
+      "loss": 0.1482,
+      "step": 2649
+    },
+    {
+      "epoch": 0.5970653673923452,
+      "grad_norm": 0.4758845188227502,
+      "learning_rate": 1.909505517324825e-06,
+      "loss": 0.1507,
+      "step": 2650
+    },
+    {
+      "epoch": 0.5972906750781536,
+      "grad_norm": 0.46282831309817246,
+      "learning_rate": 1.9076954932904e-06,
+      "loss": 0.1339,
+      "step": 2651
+    },
+    {
+      "epoch": 0.597515982763962,
+      "grad_norm": 0.47494545960669815,
+      "learning_rate": 1.905885798140591e-06,
+      "loss": 0.1451,
+      "step": 2652
+    },
+    {
+      "epoch": 0.5977412904497704,
+      "grad_norm": 0.4559165581505678,
+      "learning_rate": 1.9040764328802523e-06,
+      "loss": 0.1378,
+      "step": 2653
+    },
+    {
+      "epoch": 0.597966598135579,
+      "grad_norm": 0.4564783058150478,
+      "learning_rate": 1.9022673985140585e-06,
+      "loss": 0.1356,
+      "step": 2654
+    },
+    {
+      "epoch": 0.5981919058213874,
+      "grad_norm": 0.4860139197568067,
+      "learning_rate": 1.9004586960464993e-06,
+      "loss": 0.1581,
+      "step": 2655
+    },
+    {
+      "epoch": 0.5984172135071958,
+      "grad_norm": 0.49231192056833295,
+      "learning_rate": 1.8986503264818785e-06,
+      "loss": 0.149,
+      "step": 2656
+    },
+    {
+      "epoch": 0.5986425211930042,
+      "grad_norm": 0.46235777577164483,
+      "learning_rate": 1.8968422908243156e-06,
+      "loss": 0.1507,
+      "step": 2657
+    },
+    {
+      "epoch": 0.5988678288788126,
+      "grad_norm": 0.4574777606001307,
+      "learning_rate": 1.895034590077747e-06,
+      "loss": 0.1447,
+      "step": 2658
+    },
+    {
+      "epoch": 0.599093136564621,
+      "grad_norm": 0.518939972275897,
+      "learning_rate": 1.8932272252459213e-06,
+      "loss": 0.1661,
+      "step": 2659
+    },
+    {
+      "epoch": 0.5993184442504295,
+      "grad_norm": 0.49871042299630464,
+      "learning_rate": 1.8914201973324004e-06,
+      "loss": 0.1624,
+      "step": 2660
+    },
+    {
+      "epoch": 0.5995437519362379,
+      "grad_norm": 0.44078841524551204,
+      "learning_rate": 1.88961350734056e-06,
+      "loss": 0.1299,
+      "step": 2661
+    },
+    {
+      "epoch": 0.5997690596220464,
+      "grad_norm": 0.4788882586858163,
+      "learning_rate": 1.8878071562735873e-06,
+      "loss": 0.1593,
+      "step": 2662
+    },
+    {
+      "epoch": 0.5999943673078548,
+      "grad_norm": 0.4883771425411433,
+      "learning_rate": 1.8860011451344811e-06,
+      "loss": 0.153,
+      "step": 2663
+    },
+    {
+      "epoch": 0.6002196749936632,
+      "grad_norm": 0.4863938817343778,
+      "learning_rate": 1.8841954749260535e-06,
+      "loss": 0.1536,
+      "step": 2664
+    },
+    {
+      "epoch": 0.6004449826794717,
+      "grad_norm": 0.4956010517285103,
+      "learning_rate": 1.8823901466509253e-06,
+      "loss": 0.16,
+      "step": 2665
+    },
+    {
+      "epoch": 0.6006702903652801,
+      "grad_norm": 0.48987831491655476,
+      "learning_rate": 1.8805851613115278e-06,
+      "loss": 0.1556,
+      "step": 2666
+    },
+    {
+      "epoch": 0.6008955980510885,
+      "grad_norm": 0.45947037046050837,
+      "learning_rate": 1.878780519910102e-06,
+      "loss": 0.1343,
+      "step": 2667
+    },
+    {
+      "epoch": 0.6011209057368969,
+      "grad_norm": 0.47137603246849347,
+      "learning_rate": 1.8769762234486982e-06,
+      "loss": 0.1396,
+      "step": 2668
+    },
+    {
+      "epoch": 0.6013462134227053,
+      "grad_norm": 0.4610954891970303,
+      "learning_rate": 1.8751722729291747e-06,
+      "loss": 0.1509,
+      "step": 2669
+    },
+    {
+      "epoch": 0.6015715211085139,
+      "grad_norm": 0.45814689046776785,
+      "learning_rate": 1.8733686693531986e-06,
+      "loss": 0.1342,
+      "step": 2670
+    },
+    {
+      "epoch": 0.6017968287943223,
+      "grad_norm": 0.46749668348657086,
+      "learning_rate": 1.8715654137222434e-06,
+      "loss": 0.1385,
+      "step": 2671
+    },
+    {
+      "epoch": 0.6020221364801307,
+      "grad_norm": 0.478938533355445,
+      "learning_rate": 1.8697625070375893e-06,
+      "loss": 0.1519,
+      "step": 2672
+    },
+    {
+      "epoch": 0.6022474441659391,
+      "grad_norm": 0.47131761493341146,
+      "learning_rate": 1.8679599503003246e-06,
+      "loss": 0.1392,
+      "step": 2673
+    },
+    {
+      "epoch": 0.6024727518517475,
+      "grad_norm": 0.4492544891627898,
+      "learning_rate": 1.8661577445113399e-06,
+      "loss": 0.146,
+      "step": 2674
+    },
+    {
+      "epoch": 0.602698059537556,
+      "grad_norm": 0.49385705364348237,
+      "learning_rate": 1.8643558906713344e-06,
+      "loss": 0.1526,
+      "step": 2675
+    },
+    {
+      "epoch": 0.6029233672233644,
+      "grad_norm": 0.45914439654005745,
+      "learning_rate": 1.8625543897808094e-06,
+      "loss": 0.1371,
+      "step": 2676
+    },
+    {
+      "epoch": 0.6031486749091728,
+      "grad_norm": 0.4745967840339945,
+      "learning_rate": 1.8607532428400714e-06,
+      "loss": 0.146,
+      "step": 2677
+    },
+    {
+      "epoch": 0.6033739825949813,
+      "grad_norm": 0.470332402788067,
+      "learning_rate": 1.8589524508492308e-06,
+      "loss": 0.1597,
+      "step": 2678
+    },
+    {
+      "epoch": 0.6035992902807897,
+      "grad_norm": 0.46903431308321003,
+      "learning_rate": 1.8571520148081992e-06,
+      "loss": 0.1478,
+      "step": 2679
+    },
+    {
+      "epoch": 0.6038245979665982,
+      "grad_norm": 0.4722535666341083,
+      "learning_rate": 1.8553519357166927e-06,
+      "loss": 0.1512,
+      "step": 2680
+    },
+    {
+      "epoch": 0.6040499056524066,
+      "grad_norm": 0.47488229814225896,
+      "learning_rate": 1.853552214574228e-06,
+      "loss": 0.1517,
+      "step": 2681
+    },
+    {
+      "epoch": 0.604275213338215,
+      "grad_norm": 0.44122280475937337,
+      "learning_rate": 1.8517528523801226e-06,
+      "loss": 0.1248,
+      "step": 2682
+    },
+    {
+      "epoch": 0.6045005210240234,
+      "grad_norm": 0.4828760618283871,
+      "learning_rate": 1.8499538501334955e-06,
+      "loss": 0.1461,
+      "step": 2683
+    },
+    {
+      "epoch": 0.6047258287098318,
+      "grad_norm": 0.5094872405551787,
+      "learning_rate": 1.8481552088332656e-06,
+      "loss": 0.1521,
+      "step": 2684
+    },
+    {
+      "epoch": 0.6049511363956402,
+      "grad_norm": 0.4967269024341413,
+      "learning_rate": 1.8463569294781509e-06,
+      "loss": 0.1619,
+      "step": 2685
+    },
+    {
+      "epoch": 0.6051764440814488,
+      "grad_norm": 0.4728795946203999,
+      "learning_rate": 1.84455901306667e-06,
+      "loss": 0.1457,
+      "step": 2686
+    },
+    {
+      "epoch": 0.6054017517672572,
+      "grad_norm": 0.4499141690398027,
+      "learning_rate": 1.842761460597138e-06,
+      "loss": 0.1403,
+      "step": 2687
+    },
+    {
+      "epoch": 0.6056270594530656,
+      "grad_norm": 0.4909515705165402,
+      "learning_rate": 1.8409642730676692e-06,
+      "loss": 0.1482,
+      "step": 2688
+    },
+    {
+      "epoch": 0.605852367138874,
+      "grad_norm": 0.4877009904812424,
+      "learning_rate": 1.8391674514761745e-06,
+      "loss": 0.1478,
+      "step": 2689
+    },
+    {
+      "epoch": 0.6060776748246824,
+      "grad_norm": 0.4750945191900288,
+      "learning_rate": 1.8373709968203624e-06,
+      "loss": 0.1523,
+      "step": 2690
+    },
+    {
+      "epoch": 0.6063029825104909,
+      "grad_norm": 0.4508502573013392,
+      "learning_rate": 1.8355749100977371e-06,
+      "loss": 0.1374,
+      "step": 2691
+    },
+    {
+      "epoch": 0.6065282901962993,
+      "grad_norm": 0.45921109689505535,
+      "learning_rate": 1.8337791923055983e-06,
+      "loss": 0.1427,
+      "step": 2692
+    },
+    {
+      "epoch": 0.6067535978821077,
+      "grad_norm": 0.4453942067397089,
+      "learning_rate": 1.8319838444410412e-06,
+      "loss": 0.1296,
+      "step": 2693
+    },
+    {
+      "epoch": 0.6069789055679162,
+      "grad_norm": 0.45520091195224116,
+      "learning_rate": 1.8301888675009554e-06,
+      "loss": 0.1401,
+      "step": 2694
+    },
+    {
+      "epoch": 0.6072042132537246,
+      "grad_norm": 0.5085833788533062,
+      "learning_rate": 1.8283942624820247e-06,
+      "loss": 0.1666,
+      "step": 2695
+    },
+    {
+      "epoch": 0.6074295209395331,
+      "grad_norm": 0.46436748594980676,
+      "learning_rate": 1.8266000303807272e-06,
+      "loss": 0.142,
+      "step": 2696
+    },
+    {
+      "epoch": 0.6076548286253415,
+      "grad_norm": 0.46302456475407644,
+      "learning_rate": 1.8248061721933325e-06,
+      "loss": 0.1397,
+      "step": 2697
+    },
+    {
+      "epoch": 0.6078801363111499,
+      "grad_norm": 0.49167502175273,
+      "learning_rate": 1.8230126889159027e-06,
+      "loss": 0.1543,
+      "step": 2698
+    },
+    {
+      "epoch": 0.6081054439969583,
+      "grad_norm": 0.48436061466263614,
+      "learning_rate": 1.8212195815442934e-06,
+      "loss": 0.1502,
+      "step": 2699
+    },
+    {
+      "epoch": 0.6083307516827667,
+      "grad_norm": 0.46125404579447626,
+      "learning_rate": 1.8194268510741493e-06,
+      "loss": 0.1372,
+      "step": 2700
+    },
+    {
+      "epoch": 0.6085560593685753,
+      "grad_norm": 0.515519249077565,
+      "learning_rate": 1.8176344985009064e-06,
+      "loss": 0.1482,
+      "step": 2701
+    },
+    {
+      "epoch": 0.6087813670543837,
+      "grad_norm": 0.49611842782376675,
+      "learning_rate": 1.8158425248197931e-06,
+      "loss": 0.1566,
+      "step": 2702
+    },
+    {
+      "epoch": 0.6090066747401921,
+      "grad_norm": 0.4544133455659541,
+      "learning_rate": 1.8140509310258238e-06,
+      "loss": 0.1419,
+      "step": 2703
+    },
+    {
+      "epoch": 0.6092319824260005,
+      "grad_norm": 0.4897278959152588,
+      "learning_rate": 1.812259718113805e-06,
+      "loss": 0.1553,
+      "step": 2704
+    },
+    {
+      "epoch": 0.6094572901118089,
+      "grad_norm": 0.4550888635947903,
+      "learning_rate": 1.8104688870783296e-06,
+      "loss": 0.1421,
+      "step": 2705
+    },
+    {
+      "epoch": 0.6096825977976174,
+      "grad_norm": 0.5439820043723024,
+      "learning_rate": 1.8086784389137796e-06,
+      "loss": 0.1675,
+      "step": 2706
+    },
+    {
+      "epoch": 0.6099079054834258,
+      "grad_norm": 0.469320115995091,
+      "learning_rate": 1.806888374614324e-06,
+      "loss": 0.1489,
+      "step": 2707
+    },
+    {
+      "epoch": 0.6101332131692342,
+      "grad_norm": 0.4764614698414576,
+      "learning_rate": 1.8050986951739201e-06,
+      "loss": 0.1529,
+      "step": 2708
+    },
+    {
+      "epoch": 0.6103585208550427,
+      "grad_norm": 0.4683849207784724,
+      "learning_rate": 1.8033094015863082e-06,
+      "loss": 0.1447,
+      "step": 2709
+    },
+    {
+      "epoch": 0.6105838285408511,
+      "grad_norm": 0.47474751534392706,
+      "learning_rate": 1.8015204948450166e-06,
+      "loss": 0.142,
+      "step": 2710
+    },
+    {
+      "epoch": 0.6108091362266596,
+      "grad_norm": 0.4394569635606037,
+      "learning_rate": 1.7997319759433596e-06,
+      "loss": 0.1418,
+      "step": 2711
+    },
+    {
+      "epoch": 0.611034443912468,
+      "grad_norm": 0.48160970529582176,
+      "learning_rate": 1.7979438458744343e-06,
+      "loss": 0.1532,
+      "step": 2712
+    },
+    {
+      "epoch": 0.6112597515982764,
+      "grad_norm": 0.4645507221905752,
+      "learning_rate": 1.7961561056311234e-06,
+      "loss": 0.1527,
+      "step": 2713
+    },
+    {
+      "epoch": 0.6114850592840848,
+      "grad_norm": 0.46326252743568164,
+      "learning_rate": 1.7943687562060919e-06,
+      "loss": 0.1552,
+      "step": 2714
+    },
+    {
+      "epoch": 0.6117103669698932,
+      "grad_norm": 0.4860638067814299,
+      "learning_rate": 1.792581798591788e-06,
+      "loss": 0.1563,
+      "step": 2715
+    },
+    {
+      "epoch": 0.6119356746557016,
+      "grad_norm": 0.4748546487696476,
+      "learning_rate": 1.7907952337804429e-06,
+      "loss": 0.1459,
+      "step": 2716
+    },
+    {
+      "epoch": 0.6121609823415102,
+      "grad_norm": 0.4936458628920855,
+      "learning_rate": 1.7890090627640699e-06,
+      "loss": 0.1623,
+      "step": 2717
+    },
+    {
+      "epoch": 0.6123862900273186,
+      "grad_norm": 0.47973768416221685,
+      "learning_rate": 1.787223286534463e-06,
+      "loss": 0.1557,
+      "step": 2718
+    },
+    {
+      "epoch": 0.612611597713127,
+      "grad_norm": 0.5113779313428518,
+      "learning_rate": 1.785437906083197e-06,
+      "loss": 0.156,
+      "step": 2719
+    },
+    {
+      "epoch": 0.6128369053989354,
+      "grad_norm": 0.4957171390447654,
+      "learning_rate": 1.783652922401627e-06,
+      "loss": 0.1584,
+      "step": 2720
+    },
+    {
+      "epoch": 0.6130622130847438,
+      "grad_norm": 0.4508990357285264,
+      "learning_rate": 1.7818683364808883e-06,
+      "loss": 0.1393,
+      "step": 2721
+    },
+    {
+      "epoch": 0.6132875207705523,
+      "grad_norm": 0.46535715982732545,
+      "learning_rate": 1.7800841493118942e-06,
+      "loss": 0.1433,
+      "step": 2722
+    },
+    {
+      "epoch": 0.6135128284563607,
+      "grad_norm": 0.48346208257731116,
+      "learning_rate": 1.7783003618853384e-06,
+      "loss": 0.1511,
+      "step": 2723
+    },
+    {
+      "epoch": 0.6137381361421691,
+      "grad_norm": 0.4838514534977142,
+      "learning_rate": 1.776516975191691e-06,
+      "loss": 0.147,
+      "step": 2724
+    },
+    {
+      "epoch": 0.6139634438279776,
+      "grad_norm": 0.4768191513891604,
+      "learning_rate": 1.7747339902212e-06,
+      "loss": 0.1378,
+      "step": 2725
+    },
+    {
+      "epoch": 0.614188751513786,
+      "grad_norm": 0.4575459272190301,
+      "learning_rate": 1.7729514079638915e-06,
+      "loss": 0.1375,
+      "step": 2726
+    },
+    {
+      "epoch": 0.6144140591995945,
+      "grad_norm": 0.47887891978460684,
+      "learning_rate": 1.7711692294095654e-06,
+      "loss": 0.1575,
+      "step": 2727
+    },
+    {
+      "epoch": 0.6146393668854029,
+      "grad_norm": 0.5050928760519903,
+      "learning_rate": 1.7693874555477996e-06,
+      "loss": 0.1607,
+      "step": 2728
+    },
+    {
+      "epoch": 0.6148646745712113,
+      "grad_norm": 0.4787168044851947,
+      "learning_rate": 1.7676060873679473e-06,
+      "loss": 0.1469,
+      "step": 2729
+    },
+    {
+      "epoch": 0.6150899822570197,
+      "grad_norm": 0.48281978453942287,
+      "learning_rate": 1.7658251258591352e-06,
+      "loss": 0.1522,
+      "step": 2730
+    },
+    {
+      "epoch": 0.6153152899428281,
+      "grad_norm": 0.5109361705865251,
+      "learning_rate": 1.764044572010265e-06,
+      "loss": 0.1529,
+      "step": 2731
+    },
+    {
+      "epoch": 0.6155405976286366,
+      "grad_norm": 0.49247695655010393,
+      "learning_rate": 1.7622644268100116e-06,
+      "loss": 0.1634,
+      "step": 2732
+    },
+    {
+      "epoch": 0.6157659053144451,
+      "grad_norm": 0.5121770325492478,
+      "learning_rate": 1.7604846912468243e-06,
+      "loss": 0.1658,
+      "step": 2733
+    },
+    {
+      "epoch": 0.6159912130002535,
+      "grad_norm": 0.4749585350535519,
+      "learning_rate": 1.7587053663089233e-06,
+      "loss": 0.1457,
+      "step": 2734
+    },
+    {
+      "epoch": 0.6162165206860619,
+      "grad_norm": 0.4709334071762967,
+      "learning_rate": 1.7569264529843009e-06,
+      "loss": 0.1529,
+      "step": 2735
+    },
+    {
+      "epoch": 0.6164418283718703,
+      "grad_norm": 0.47968940655553516,
+      "learning_rate": 1.755147952260722e-06,
+      "loss": 0.1485,
+      "step": 2736
+    },
+    {
+      "epoch": 0.6166671360576788,
+      "grad_norm": 0.48261957330261573,
+      "learning_rate": 1.753369865125722e-06,
+      "loss": 0.1581,
+      "step": 2737
+    },
+    {
+      "epoch": 0.6168924437434872,
+      "grad_norm": 0.4795332092297908,
+      "learning_rate": 1.7515921925666053e-06,
+      "loss": 0.1568,
+      "step": 2738
+    },
+    {
+      "epoch": 0.6171177514292956,
+      "grad_norm": 0.4717642359972822,
+      "learning_rate": 1.749814935570448e-06,
+      "loss": 0.1558,
+      "step": 2739
+    },
+    {
+      "epoch": 0.617343059115104,
+      "grad_norm": 0.4840128981249241,
+      "learning_rate": 1.748038095124095e-06,
+      "loss": 0.1455,
+      "step": 2740
+    },
+    {
+      "epoch": 0.6175683668009125,
+      "grad_norm": 0.465278403538817,
+      "learning_rate": 1.746261672214159e-06,
+      "loss": 0.1433,
+      "step": 2741
+    },
+    {
+      "epoch": 0.617793674486721,
+      "grad_norm": 0.48162713449264705,
+      "learning_rate": 1.7444856678270218e-06,
+      "loss": 0.1569,
+      "step": 2742
+    },
+    {
+      "epoch": 0.6180189821725294,
+      "grad_norm": 0.4832915581984926,
+      "learning_rate": 1.7427100829488325e-06,
+      "loss": 0.1487,
+      "step": 2743
+    },
+    {
+      "epoch": 0.6182442898583378,
+      "grad_norm": 0.4805571768792867,
+      "learning_rate": 1.7409349185655067e-06,
+      "loss": 0.1532,
+      "step": 2744
+    },
+    {
+      "epoch": 0.6184695975441462,
+      "grad_norm": 0.45648256848154184,
+      "learning_rate": 1.739160175662727e-06,
+      "loss": 0.1424,
+      "step": 2745
+    },
+    {
+      "epoch": 0.6186949052299546,
+      "grad_norm": 0.4588732490898067,
+      "learning_rate": 1.7373858552259421e-06,
+      "loss": 0.1522,
+      "step": 2746
+    },
+    {
+      "epoch": 0.618920212915763,
+      "grad_norm": 0.47375321309353036,
+      "learning_rate": 1.7356119582403663e-06,
+      "loss": 0.1474,
+      "step": 2747
+    },
+    {
+      "epoch": 0.6191455206015716,
+      "grad_norm": 0.5007531820480831,
+      "learning_rate": 1.733838485690978e-06,
+      "loss": 0.1501,
+      "step": 2748
+    },
+    {
+      "epoch": 0.61937082828738,
+      "grad_norm": 0.480719639702643,
+      "learning_rate": 1.7320654385625208e-06,
+      "loss": 0.141,
+      "step": 2749
+    },
+    {
+      "epoch": 0.6195961359731884,
+      "grad_norm": 0.4587097913676663,
+      "learning_rate": 1.7302928178395018e-06,
+      "loss": 0.1388,
+      "step": 2750
+    },
+    {
+      "epoch": 0.6198214436589968,
+      "grad_norm": 0.4958585682683732,
+      "learning_rate": 1.7285206245061908e-06,
+      "loss": 0.1503,
+      "step": 2751
+    },
+    {
+      "epoch": 0.6200467513448052,
+      "grad_norm": 0.4550933127673,
+      "learning_rate": 1.726748859546621e-06,
+      "loss": 0.1391,
+      "step": 2752
+    },
+    {
+      "epoch": 0.6202720590306137,
+      "grad_norm": 0.44816444530545974,
+      "learning_rate": 1.7249775239445875e-06,
+      "loss": 0.1307,
+      "step": 2753
+    },
+    {
+      "epoch": 0.6204973667164221,
+      "grad_norm": 0.4431721028100357,
+      "learning_rate": 1.723206618683646e-06,
+      "loss": 0.1331,
+      "step": 2754
+    },
+    {
+      "epoch": 0.6207226744022305,
+      "grad_norm": 0.511980981222847,
+      "learning_rate": 1.7214361447471156e-06,
+      "loss": 0.1493,
+      "step": 2755
+    },
+    {
+      "epoch": 0.620947982088039,
+      "grad_norm": 0.47289755552146395,
+      "learning_rate": 1.7196661031180738e-06,
+      "loss": 0.1585,
+      "step": 2756
+    },
+    {
+      "epoch": 0.6211732897738474,
+      "grad_norm": 0.47535833672053196,
+      "learning_rate": 1.7178964947793591e-06,
+      "loss": 0.1437,
+      "step": 2757
+    },
+    {
+      "epoch": 0.6213985974596559,
+      "grad_norm": 0.48646685714584237,
+      "learning_rate": 1.716127320713568e-06,
+      "loss": 0.146,
+      "step": 2758
+    },
+    {
+      "epoch": 0.6216239051454643,
+      "grad_norm": 0.4634324355114312,
+      "learning_rate": 1.7143585819030583e-06,
+      "loss": 0.143,
+      "step": 2759
+    },
+    {
+      "epoch": 0.6218492128312727,
+      "grad_norm": 0.4576000720937714,
+      "learning_rate": 1.7125902793299434e-06,
+      "loss": 0.1371,
+      "step": 2760
+    },
+    {
+      "epoch": 0.6220745205170811,
+      "grad_norm": 0.4438078503440971,
+      "learning_rate": 1.7108224139760982e-06,
+      "loss": 0.1249,
+      "step": 2761
+    },
+    {
+      "epoch": 0.6222998282028895,
+      "grad_norm": 0.4471311093079337,
+      "learning_rate": 1.7090549868231492e-06,
+      "loss": 0.1305,
+      "step": 2762
+    },
+    {
+      "epoch": 0.622525135888698,
+      "grad_norm": 0.4808441910375895,
+      "learning_rate": 1.707287998852485e-06,
+      "loss": 0.1548,
+      "step": 2763
+    },
+    {
+      "epoch": 0.6227504435745065,
+      "grad_norm": 0.4401254562234489,
+      "learning_rate": 1.7055214510452462e-06,
+      "loss": 0.1387,
+      "step": 2764
+    },
+    {
+      "epoch": 0.6229757512603149,
+      "grad_norm": 0.4879263554502538,
+      "learning_rate": 1.7037553443823332e-06,
+      "loss": 0.1465,
+      "step": 2765
+    },
+    {
+      "epoch": 0.6232010589461233,
+      "grad_norm": 0.4472646910236901,
+      "learning_rate": 1.7019896798443984e-06,
+      "loss": 0.1373,
+      "step": 2766
+    },
+    {
+      "epoch": 0.6234263666319317,
+      "grad_norm": 0.48360400598865216,
+      "learning_rate": 1.700224458411849e-06,
+      "loss": 0.1559,
+      "step": 2767
+    },
+    {
+      "epoch": 0.6236516743177402,
+      "grad_norm": 0.4830733201104498,
+      "learning_rate": 1.6984596810648475e-06,
+      "loss": 0.1579,
+      "step": 2768
+    },
+    {
+      "epoch": 0.6238769820035486,
+      "grad_norm": 0.48335608809827474,
+      "learning_rate": 1.6966953487833078e-06,
+      "loss": 0.1596,
+      "step": 2769
+    },
+    {
+      "epoch": 0.624102289689357,
+      "grad_norm": 0.4823123539206247,
+      "learning_rate": 1.6949314625468985e-06,
+      "loss": 0.1513,
+      "step": 2770
+    },
+    {
+      "epoch": 0.6243275973751654,
+      "grad_norm": 0.4938808073292687,
+      "learning_rate": 1.6931680233350404e-06,
+      "loss": 0.151,
+      "step": 2771
+    },
+    {
+      "epoch": 0.6245529050609739,
+      "grad_norm": 0.47738363812247797,
+      "learning_rate": 1.6914050321269049e-06,
+      "loss": 0.1473,
+      "step": 2772
+    },
+    {
+      "epoch": 0.6247782127467824,
+      "grad_norm": 0.4889654971871984,
+      "learning_rate": 1.6896424899014158e-06,
+      "loss": 0.1413,
+      "step": 2773
+    },
+    {
+      "epoch": 0.6250035204325908,
+      "grad_norm": 0.4903261247950262,
+      "learning_rate": 1.6878803976372465e-06,
+      "loss": 0.1663,
+      "step": 2774
+    },
+    {
+      "epoch": 0.6252288281183992,
+      "grad_norm": 0.48272666070040504,
+      "learning_rate": 1.6861187563128217e-06,
+      "loss": 0.1535,
+      "step": 2775
+    },
+    {
+      "epoch": 0.6254541358042076,
+      "grad_norm": 0.46179189709660806,
+      "learning_rate": 1.6843575669063142e-06,
+      "loss": 0.1375,
+      "step": 2776
+    },
+    {
+      "epoch": 0.625679443490016,
+      "grad_norm": 0.46497370122587345,
+      "learning_rate": 1.682596830395648e-06,
+      "loss": 0.1407,
+      "step": 2777
+    },
+    {
+      "epoch": 0.6259047511758244,
+      "grad_norm": 0.4702044055000179,
+      "learning_rate": 1.6808365477584953e-06,
+      "loss": 0.1534,
+      "step": 2778
+    },
+    {
+      "epoch": 0.6261300588616329,
+      "grad_norm": 0.5071413974321057,
+      "learning_rate": 1.6790767199722724e-06,
+      "loss": 0.1626,
+      "step": 2779
+    },
+    {
+      "epoch": 0.6263553665474414,
+      "grad_norm": 0.4805694941313356,
+      "learning_rate": 1.6773173480141487e-06,
+      "loss": 0.1482,
+      "step": 2780
+    },
+    {
+      "epoch": 0.6265806742332498,
+      "grad_norm": 0.46670874376304294,
+      "learning_rate": 1.6755584328610364e-06,
+      "loss": 0.1446,
+      "step": 2781
+    },
+    {
+      "epoch": 0.6268059819190582,
+      "grad_norm": 0.48126522962899204,
+      "learning_rate": 1.6737999754895965e-06,
+      "loss": 0.1433,
+      "step": 2782
+    },
+    {
+      "epoch": 0.6270312896048666,
+      "grad_norm": 0.47189768674803717,
+      "learning_rate": 1.6720419768762343e-06,
+      "loss": 0.1515,
+      "step": 2783
+    },
+    {
+      "epoch": 0.6272565972906751,
+      "grad_norm": 0.4721444472897269,
+      "learning_rate": 1.6702844379971012e-06,
+      "loss": 0.1427,
+      "step": 2784
+    },
+    {
+      "epoch": 0.6274819049764835,
+      "grad_norm": 0.4876297281006269,
+      "learning_rate": 1.668527359828092e-06,
+      "loss": 0.1504,
+      "step": 2785
+    },
+    {
+      "epoch": 0.6277072126622919,
+      "grad_norm": 0.4767240074099445,
+      "learning_rate": 1.6667707433448482e-06,
+      "loss": 0.1532,
+      "step": 2786
+    },
+    {
+      "epoch": 0.6279325203481003,
+      "grad_norm": 0.47858125114578143,
+      "learning_rate": 1.6650145895227532e-06,
+      "loss": 0.1551,
+      "step": 2787
+    },
+    {
+      "epoch": 0.6281578280339088,
+      "grad_norm": 0.4859216333502786,
+      "learning_rate": 1.663258899336933e-06,
+      "loss": 0.159,
+      "step": 2788
+    },
+    {
+      "epoch": 0.6283831357197173,
+      "grad_norm": 0.46248161988504,
+      "learning_rate": 1.6615036737622574e-06,
+      "loss": 0.1421,
+      "step": 2789
+    },
+    {
+      "epoch": 0.6286084434055257,
+      "grad_norm": 0.5055365786287488,
+      "learning_rate": 1.6597489137733377e-06,
+      "loss": 0.1597,
+      "step": 2790
+    },
+    {
+      "epoch": 0.6288337510913341,
+      "grad_norm": 0.4743015397097211,
+      "learning_rate": 1.6579946203445269e-06,
+      "loss": 0.1489,
+      "step": 2791
+    },
+    {
+      "epoch": 0.6290590587771425,
+      "grad_norm": 0.46321209388814183,
+      "learning_rate": 1.6562407944499175e-06,
+      "loss": 0.1405,
+      "step": 2792
+    },
+    {
+      "epoch": 0.6292843664629509,
+      "grad_norm": 0.47427596876506406,
+      "learning_rate": 1.6544874370633456e-06,
+      "loss": 0.1389,
+      "step": 2793
+    },
+    {
+      "epoch": 0.6295096741487594,
+      "grad_norm": 0.48628130797575075,
+      "learning_rate": 1.652734549158384e-06,
+      "loss": 0.1556,
+      "step": 2794
+    },
+    {
+      "epoch": 0.6297349818345679,
+      "grad_norm": 0.48296863875610585,
+      "learning_rate": 1.6509821317083466e-06,
+      "loss": 0.1471,
+      "step": 2795
+    },
+    {
+      "epoch": 0.6299602895203763,
+      "grad_norm": 0.44435202435685495,
+      "learning_rate": 1.6492301856862855e-06,
+      "loss": 0.1316,
+      "step": 2796
+    },
+    {
+      "epoch": 0.6301855972061847,
+      "grad_norm": 0.4522766714459569,
+      "learning_rate": 1.6474787120649903e-06,
+      "loss": 0.1491,
+      "step": 2797
+    },
+    {
+      "epoch": 0.6304109048919931,
+      "grad_norm": 0.476592061403675,
+      "learning_rate": 1.6457277118169893e-06,
+      "loss": 0.1495,
+      "step": 2798
+    },
+    {
+      "epoch": 0.6306362125778016,
+      "grad_norm": 0.49291809206381026,
+      "learning_rate": 1.6439771859145476e-06,
+      "loss": 0.1477,
+      "step": 2799
+    },
+    {
+      "epoch": 0.63086152026361,
+      "grad_norm": 0.4712511414943249,
+      "learning_rate": 1.6422271353296675e-06,
+      "loss": 0.1457,
+      "step": 2800
+    },
+    {
+      "epoch": 0.6310868279494184,
+      "grad_norm": 0.5061894262618007,
+      "learning_rate": 1.640477561034086e-06,
+      "loss": 0.1494,
+      "step": 2801
+    },
+    {
+      "epoch": 0.6313121356352268,
+      "grad_norm": 0.45287458848331624,
+      "learning_rate": 1.6387284639992773e-06,
+      "loss": 0.139,
+      "step": 2802
+    },
+    {
+      "epoch": 0.6315374433210353,
+      "grad_norm": 0.4723200403220778,
+      "learning_rate": 1.6369798451964496e-06,
+      "loss": 0.1472,
+      "step": 2803
+    },
+    {
+      "epoch": 0.6317627510068438,
+      "grad_norm": 0.5103378496664759,
+      "learning_rate": 1.6352317055965458e-06,
+      "loss": 0.159,
+      "step": 2804
+    },
+    {
+      "epoch": 0.6319880586926522,
+      "grad_norm": 0.480394644135576,
+      "learning_rate": 1.6334840461702422e-06,
+      "loss": 0.1516,
+      "step": 2805
+    },
+    {
+      "epoch": 0.6322133663784606,
+      "grad_norm": 0.47523540206244513,
+      "learning_rate": 1.6317368678879497e-06,
+      "loss": 0.1389,
+      "step": 2806
+    },
+    {
+      "epoch": 0.632438674064269,
+      "grad_norm": 0.4558136082094893,
+      "learning_rate": 1.6299901717198102e-06,
+      "loss": 0.1382,
+      "step": 2807
+    },
+    {
+      "epoch": 0.6326639817500774,
+      "grad_norm": 0.48295797707491017,
+      "learning_rate": 1.6282439586356999e-06,
+      "loss": 0.151,
+      "step": 2808
+    },
+    {
+      "epoch": 0.6328892894358858,
+      "grad_norm": 0.4872663984317607,
+      "learning_rate": 1.6264982296052256e-06,
+      "loss": 0.1483,
+      "step": 2809
+    },
+    {
+      "epoch": 0.6331145971216943,
+      "grad_norm": 0.44741252541960735,
+      "learning_rate": 1.6247529855977256e-06,
+      "loss": 0.1342,
+      "step": 2810
+    },
+    {
+      "epoch": 0.6333399048075028,
+      "grad_norm": 0.4985585348392438,
+      "learning_rate": 1.6230082275822687e-06,
+      "loss": 0.1531,
+      "step": 2811
+    },
+    {
+      "epoch": 0.6335652124933112,
+      "grad_norm": 0.4638990385262181,
+      "learning_rate": 1.6212639565276538e-06,
+      "loss": 0.136,
+      "step": 2812
+    },
+    {
+      "epoch": 0.6337905201791196,
+      "grad_norm": 0.4650018045050846,
+      "learning_rate": 1.6195201734024096e-06,
+      "loss": 0.1381,
+      "step": 2813
+    },
+    {
+      "epoch": 0.634015827864928,
+      "grad_norm": 0.4965714175065691,
+      "learning_rate": 1.6177768791747957e-06,
+      "loss": 0.1496,
+      "step": 2814
+    },
+    {
+      "epoch": 0.6342411355507365,
+      "grad_norm": 0.47979460480702335,
+      "learning_rate": 1.6160340748127959e-06,
+      "loss": 0.1513,
+      "step": 2815
+    },
+    {
+      "epoch": 0.6344664432365449,
+      "grad_norm": 0.442911411378338,
+      "learning_rate": 1.6142917612841252e-06,
+      "loss": 0.1299,
+      "step": 2816
+    },
+    {
+      "epoch": 0.6346917509223533,
+      "grad_norm": 0.48375134793054897,
+      "learning_rate": 1.612549939556225e-06,
+      "loss": 0.1397,
+      "step": 2817
+    },
+    {
+      "epoch": 0.6349170586081617,
+      "grad_norm": 0.4646454924086842,
+      "learning_rate": 1.610808610596265e-06,
+      "loss": 0.1407,
+      "step": 2818
+    },
+    {
+      "epoch": 0.6351423662939703,
+      "grad_norm": 0.47665971541619717,
+      "learning_rate": 1.6090677753711403e-06,
+      "loss": 0.1385,
+      "step": 2819
+    },
+    {
+      "epoch": 0.6353676739797787,
+      "grad_norm": 0.48639642703567026,
+      "learning_rate": 1.607327434847471e-06,
+      "loss": 0.1545,
+      "step": 2820
+    },
+    {
+      "epoch": 0.6355929816655871,
+      "grad_norm": 0.5270081816426536,
+      "learning_rate": 1.6055875899916034e-06,
+      "loss": 0.1552,
+      "step": 2821
+    },
+    {
+      "epoch": 0.6358182893513955,
+      "grad_norm": 0.47067747824845496,
+      "learning_rate": 1.6038482417696095e-06,
+      "loss": 0.1516,
+      "step": 2822
+    },
+    {
+      "epoch": 0.6360435970372039,
+      "grad_norm": 0.4850765057650214,
+      "learning_rate": 1.6021093911472825e-06,
+      "loss": 0.1534,
+      "step": 2823
+    },
+    {
+      "epoch": 0.6362689047230123,
+      "grad_norm": 0.47449309561057146,
+      "learning_rate": 1.6003710390901434e-06,
+      "loss": 0.1542,
+      "step": 2824
+    },
+    {
+      "epoch": 0.6364942124088208,
+      "grad_norm": 0.5126023870225591,
+      "learning_rate": 1.5986331865634335e-06,
+      "loss": 0.1659,
+      "step": 2825
+    },
+    {
+      "epoch": 0.6367195200946292,
+      "grad_norm": 0.5052408111958812,
+      "learning_rate": 1.5968958345321178e-06,
+      "loss": 0.1526,
+      "step": 2826
+    },
+    {
+      "epoch": 0.6369448277804377,
+      "grad_norm": 0.47297868340618493,
+      "learning_rate": 1.5951589839608828e-06,
+      "loss": 0.1466,
+      "step": 2827
+    },
+    {
+      "epoch": 0.6371701354662461,
+      "grad_norm": 0.46382611744276403,
+      "learning_rate": 1.5934226358141368e-06,
+      "loss": 0.1391,
+      "step": 2828
+    },
+    {
+      "epoch": 0.6373954431520545,
+      "grad_norm": 0.49119112893116423,
+      "learning_rate": 1.5916867910560092e-06,
+      "loss": 0.1491,
+      "step": 2829
+    },
+    {
+      "epoch": 0.637620750837863,
+      "grad_norm": 0.5016831914604146,
+      "learning_rate": 1.5899514506503499e-06,
+      "loss": 0.1649,
+      "step": 2830
+    },
+    {
+      "epoch": 0.6378460585236714,
+      "grad_norm": 0.4817486742658559,
+      "learning_rate": 1.5882166155607306e-06,
+      "loss": 0.1592,
+      "step": 2831
+    },
+    {
+      "epoch": 0.6380713662094798,
+      "grad_norm": 0.46605829216160005,
+      "learning_rate": 1.5864822867504376e-06,
+      "loss": 0.1467,
+      "step": 2832
+    },
+    {
+      "epoch": 0.6382966738952882,
+      "grad_norm": 0.4804265967852996,
+      "learning_rate": 1.58474846518248e-06,
+      "loss": 0.152,
+      "step": 2833
+    },
+    {
+      "epoch": 0.6385219815810966,
+      "grad_norm": 0.4457918826886218,
+      "learning_rate": 1.5830151518195846e-06,
+      "loss": 0.1393,
+      "step": 2834
+    },
+    {
+      "epoch": 0.6387472892669052,
+      "grad_norm": 0.4545444717911307,
+      "learning_rate": 1.5812823476241962e-06,
+      "loss": 0.1403,
+      "step": 2835
+    },
+    {
+      "epoch": 0.6389725969527136,
+      "grad_norm": 0.4756372945724833,
+      "learning_rate": 1.5795500535584758e-06,
+      "loss": 0.1511,
+      "step": 2836
+    },
+    {
+      "epoch": 0.639197904638522,
+      "grad_norm": 0.4780977819516656,
+      "learning_rate": 1.5778182705843017e-06,
+      "loss": 0.1546,
+      "step": 2837
+    },
+    {
+      "epoch": 0.6394232123243304,
+      "grad_norm": 0.47229095632094537,
+      "learning_rate": 1.5760869996632685e-06,
+      "loss": 0.1507,
+      "step": 2838
+    },
+    {
+      "epoch": 0.6396485200101388,
+      "grad_norm": 0.47035365018536524,
+      "learning_rate": 1.574356241756686e-06,
+      "loss": 0.1427,
+      "step": 2839
+    },
+    {
+      "epoch": 0.6398738276959473,
+      "grad_norm": 0.48265274771659217,
+      "learning_rate": 1.572625997825581e-06,
+      "loss": 0.151,
+      "step": 2840
+    },
+    {
+      "epoch": 0.6400991353817557,
+      "grad_norm": 0.4813791272462431,
+      "learning_rate": 1.5708962688306916e-06,
+      "loss": 0.1607,
+      "step": 2841
+    },
+    {
+      "epoch": 0.6403244430675642,
+      "grad_norm": 0.4931264969520297,
+      "learning_rate": 1.5691670557324734e-06,
+      "loss": 0.1507,
+      "step": 2842
+    },
+    {
+      "epoch": 0.6405497507533726,
+      "grad_norm": 0.4891345812437931,
+      "learning_rate": 1.5674383594910931e-06,
+      "loss": 0.1528,
+      "step": 2843
+    },
+    {
+      "epoch": 0.640775058439181,
+      "grad_norm": 0.4734338447627687,
+      "learning_rate": 1.5657101810664314e-06,
+      "loss": 0.1533,
+      "step": 2844
+    },
+    {
+      "epoch": 0.6410003661249895,
+      "grad_norm": 0.4523902167413687,
+      "learning_rate": 1.5639825214180808e-06,
+      "loss": 0.1333,
+      "step": 2845
+    },
+    {
+      "epoch": 0.6412256738107979,
+      "grad_norm": 0.5081364752489811,
+      "learning_rate": 1.5622553815053476e-06,
+      "loss": 0.1551,
+      "step": 2846
+    },
+    {
+      "epoch": 0.6414509814966063,
+      "grad_norm": 0.46507190532041826,
+      "learning_rate": 1.5605287622872478e-06,
+      "loss": 0.1504,
+      "step": 2847
+    },
+    {
+      "epoch": 0.6416762891824147,
+      "grad_norm": 0.4641771875874691,
+      "learning_rate": 1.558802664722508e-06,
+      "loss": 0.1371,
+      "step": 2848
+    },
+    {
+      "epoch": 0.6419015968682231,
+      "grad_norm": 0.45632753268823106,
+      "learning_rate": 1.5570770897695672e-06,
+      "loss": 0.1445,
+      "step": 2849
+    },
+    {
+      "epoch": 0.6421269045540317,
+      "grad_norm": 0.49783070157795684,
+      "learning_rate": 1.555352038386571e-06,
+      "loss": 0.1503,
+      "step": 2850
+    },
+    {
+      "epoch": 0.6423522122398401,
+      "grad_norm": 0.46436518926277925,
+      "learning_rate": 1.5536275115313776e-06,
+      "loss": 0.1424,
+      "step": 2851
+    },
+    {
+      "epoch": 0.6425775199256485,
+      "grad_norm": 0.48717850193393475,
+      "learning_rate": 1.5519035101615518e-06,
+      "loss": 0.152,
+      "step": 2852
+    },
+    {
+      "epoch": 0.6428028276114569,
+      "grad_norm": 0.4910056281463092,
+      "learning_rate": 1.5501800352343673e-06,
+      "loss": 0.1539,
+      "step": 2853
+    },
+    {
+      "epoch": 0.6430281352972653,
+      "grad_norm": 0.5118422933613982,
+      "learning_rate": 1.5484570877068055e-06,
+      "loss": 0.155,
+      "step": 2854
+    },
+    {
+      "epoch": 0.6432534429830737,
+      "grad_norm": 0.5310399027828411,
+      "learning_rate": 1.5467346685355553e-06,
+      "loss": 0.1683,
+      "step": 2855
+    },
+    {
+      "epoch": 0.6434787506688822,
+      "grad_norm": 0.506021853271099,
+      "learning_rate": 1.5450127786770116e-06,
+      "loss": 0.1571,
+      "step": 2856
+    },
+    {
+      "epoch": 0.6437040583546906,
+      "grad_norm": 0.49174135641662325,
+      "learning_rate": 1.5432914190872757e-06,
+      "loss": 0.1518,
+      "step": 2857
+    },
+    {
+      "epoch": 0.6439293660404991,
+      "grad_norm": 0.46911119250398337,
+      "learning_rate": 1.5415705907221545e-06,
+      "loss": 0.1364,
+      "step": 2858
+    },
+    {
+      "epoch": 0.6441546737263075,
+      "grad_norm": 0.4843368641508339,
+      "learning_rate": 1.53985029453716e-06,
+      "loss": 0.1421,
+      "step": 2859
+    },
+    {
+      "epoch": 0.6443799814121159,
+      "grad_norm": 0.5137282334737647,
+      "learning_rate": 1.5381305314875084e-06,
+      "loss": 0.1565,
+      "step": 2860
+    },
+    {
+      "epoch": 0.6446052890979244,
+      "grad_norm": 0.4661382667029591,
+      "learning_rate": 1.536411302528119e-06,
+      "loss": 0.1484,
+      "step": 2861
+    },
+    {
+      "epoch": 0.6448305967837328,
+      "grad_norm": 0.4698458690541328,
+      "learning_rate": 1.5346926086136171e-06,
+      "loss": 0.1496,
+      "step": 2862
+    },
+    {
+      "epoch": 0.6450559044695412,
+      "grad_norm": 0.4734423472032637,
+      "learning_rate": 1.5329744506983292e-06,
+      "loss": 0.1532,
+      "step": 2863
+    },
+    {
+      "epoch": 0.6452812121553496,
+      "grad_norm": 0.4793175583549265,
+      "learning_rate": 1.5312568297362834e-06,
+      "loss": 0.137,
+      "step": 2864
+    },
+    {
+      "epoch": 0.645506519841158,
+      "grad_norm": 0.5120639200872323,
+      "learning_rate": 1.5295397466812115e-06,
+      "loss": 0.1515,
+      "step": 2865
+    },
+    {
+      "epoch": 0.6457318275269666,
+      "grad_norm": 0.48304485914928136,
+      "learning_rate": 1.5278232024865458e-06,
+      "loss": 0.1449,
+      "step": 2866
+    },
+    {
+      "epoch": 0.645957135212775,
+      "grad_norm": 0.478700554504123,
+      "learning_rate": 1.5261071981054183e-06,
+      "loss": 0.1512,
+      "step": 2867
+    },
+    {
+      "epoch": 0.6461824428985834,
+      "grad_norm": 0.4884664261324441,
+      "learning_rate": 1.5243917344906625e-06,
+      "loss": 0.1459,
+      "step": 2868
+    },
+    {
+      "epoch": 0.6464077505843918,
+      "grad_norm": 0.45259907738132826,
+      "learning_rate": 1.5226768125948122e-06,
+      "loss": 0.135,
+      "step": 2869
+    },
+    {
+      "epoch": 0.6466330582702002,
+      "grad_norm": 0.43908078665268413,
+      "learning_rate": 1.5209624333700985e-06,
+      "loss": 0.1315,
+      "step": 2870
+    },
+    {
+      "epoch": 0.6468583659560087,
+      "grad_norm": 0.4545357145159809,
+      "learning_rate": 1.5192485977684528e-06,
+      "loss": 0.1501,
+      "step": 2871
+    },
+    {
+      "epoch": 0.6470836736418171,
+      "grad_norm": 0.49335283447767514,
+      "learning_rate": 1.517535306741505e-06,
+      "loss": 0.1559,
+      "step": 2872
+    },
+    {
+      "epoch": 0.6473089813276255,
+      "grad_norm": 0.48260902139598805,
+      "learning_rate": 1.5158225612405808e-06,
+      "loss": 0.1402,
+      "step": 2873
+    },
+    {
+      "epoch": 0.647534289013434,
+      "grad_norm": 0.4544937631611182,
+      "learning_rate": 1.5141103622167042e-06,
+      "loss": 0.1384,
+      "step": 2874
+    },
+    {
+      "epoch": 0.6477595966992424,
+      "grad_norm": 0.4501153858395984,
+      "learning_rate": 1.512398710620595e-06,
+      "loss": 0.1427,
+      "step": 2875
+    },
+    {
+      "epoch": 0.6479849043850509,
+      "grad_norm": 0.44247881743042544,
+      "learning_rate": 1.51068760740267e-06,
+      "loss": 0.1346,
+      "step": 2876
+    },
+    {
+      "epoch": 0.6482102120708593,
+      "grad_norm": 0.4548380734877828,
+      "learning_rate": 1.508977053513041e-06,
+      "loss": 0.1371,
+      "step": 2877
+    },
+    {
+      "epoch": 0.6484355197566677,
+      "grad_norm": 0.45590988633602597,
+      "learning_rate": 1.5072670499015151e-06,
+      "loss": 0.1413,
+      "step": 2878
+    },
+    {
+      "epoch": 0.6486608274424761,
+      "grad_norm": 0.4894275363782382,
+      "learning_rate": 1.5055575975175929e-06,
+      "loss": 0.1607,
+      "step": 2879
+    },
+    {
+      "epoch": 0.6488861351282845,
+      "grad_norm": 0.5019183155077958,
+      "learning_rate": 1.5038486973104704e-06,
+      "loss": 0.1625,
+      "step": 2880
+    },
+    {
+      "epoch": 0.6491114428140929,
+      "grad_norm": 0.46667254314094453,
+      "learning_rate": 1.5021403502290354e-06,
+      "loss": 0.1344,
+      "step": 2881
+    },
+    {
+      "epoch": 0.6493367504999015,
+      "grad_norm": 0.47319219186852757,
+      "learning_rate": 1.5004325572218698e-06,
+      "loss": 0.142,
+      "step": 2882
+    },
+    {
+      "epoch": 0.6495620581857099,
+      "grad_norm": 0.4960962366132025,
+      "learning_rate": 1.4987253192372471e-06,
+      "loss": 0.157,
+      "step": 2883
+    },
+    {
+      "epoch": 0.6497873658715183,
+      "grad_norm": 0.4787973264187811,
+      "learning_rate": 1.4970186372231347e-06,
+      "loss": 0.1518,
+      "step": 2884
+    },
+    {
+      "epoch": 0.6500126735573267,
+      "grad_norm": 0.5159643516295653,
+      "learning_rate": 1.4953125121271866e-06,
+      "loss": 0.1581,
+      "step": 2885
+    },
+    {
+      "epoch": 0.6502379812431351,
+      "grad_norm": 0.4665006065685688,
+      "learning_rate": 1.493606944896751e-06,
+      "loss": 0.1404,
+      "step": 2886
+    },
+    {
+      "epoch": 0.6504632889289436,
+      "grad_norm": 0.45934804469692997,
+      "learning_rate": 1.4919019364788678e-06,
+      "loss": 0.1351,
+      "step": 2887
+    },
+    {
+      "epoch": 0.650688596614752,
+      "grad_norm": 0.4900662596708013,
+      "learning_rate": 1.490197487820263e-06,
+      "loss": 0.1548,
+      "step": 2888
+    },
+    {
+      "epoch": 0.6509139043005604,
+      "grad_norm": 0.4609436190944118,
+      "learning_rate": 1.4884935998673539e-06,
+      "loss": 0.1472,
+      "step": 2889
+    },
+    {
+      "epoch": 0.6511392119863689,
+      "grad_norm": 0.4762971414293838,
+      "learning_rate": 1.486790273566246e-06,
+      "loss": 0.1546,
+      "step": 2890
+    },
+    {
+      "epoch": 0.6513645196721773,
+      "grad_norm": 0.4556504565623819,
+      "learning_rate": 1.4850875098627326e-06,
+      "loss": 0.1453,
+      "step": 2891
+    },
+    {
+      "epoch": 0.6515898273579858,
+      "grad_norm": 0.49763192413888446,
+      "learning_rate": 1.483385309702295e-06,
+      "loss": 0.1561,
+      "step": 2892
+    },
+    {
+      "epoch": 0.6518151350437942,
+      "grad_norm": 0.4766316614497349,
+      "learning_rate": 1.4816836740301019e-06,
+      "loss": 0.1513,
+      "step": 2893
+    },
+    {
+      "epoch": 0.6520404427296026,
+      "grad_norm": 0.5184596612142391,
+      "learning_rate": 1.4799826037910082e-06,
+      "loss": 0.1699,
+      "step": 2894
+    },
+    {
+      "epoch": 0.652265750415411,
+      "grad_norm": 0.4536514939994665,
+      "learning_rate": 1.478282099929554e-06,
+      "loss": 0.1354,
+      "step": 2895
+    },
+    {
+      "epoch": 0.6524910581012194,
+      "grad_norm": 0.4974095639499098,
+      "learning_rate": 1.4765821633899663e-06,
+      "loss": 0.151,
+      "step": 2896
+    },
+    {
+      "epoch": 0.652716365787028,
+      "grad_norm": 0.4712272360337234,
+      "learning_rate": 1.4748827951161566e-06,
+      "loss": 0.1417,
+      "step": 2897
+    },
+    {
+      "epoch": 0.6529416734728364,
+      "grad_norm": 0.475826250525577,
+      "learning_rate": 1.4731839960517202e-06,
+      "loss": 0.1504,
+      "step": 2898
+    },
+    {
+      "epoch": 0.6531669811586448,
+      "grad_norm": 0.5037284144728293,
+      "learning_rate": 1.4714857671399374e-06,
+      "loss": 0.1651,
+      "step": 2899
+    },
+    {
+      "epoch": 0.6533922888444532,
+      "grad_norm": 0.45473879817414803,
+      "learning_rate": 1.4697881093237714e-06,
+      "loss": 0.144,
+      "step": 2900
+    },
+    {
+      "epoch": 0.6536175965302616,
+      "grad_norm": 0.47770176405517745,
+      "learning_rate": 1.4680910235458692e-06,
+      "loss": 0.1471,
+      "step": 2901
+    },
+    {
+      "epoch": 0.65384290421607,
+      "grad_norm": 0.4723218687376834,
+      "learning_rate": 1.4663945107485567e-06,
+      "loss": 0.1431,
+      "step": 2902
+    },
+    {
+      "epoch": 0.6540682119018785,
+      "grad_norm": 0.47416198265004644,
+      "learning_rate": 1.4646985718738466e-06,
+      "loss": 0.1491,
+      "step": 2903
+    },
+    {
+      "epoch": 0.6542935195876869,
+      "grad_norm": 0.47272637335722917,
+      "learning_rate": 1.4630032078634293e-06,
+      "loss": 0.1442,
+      "step": 2904
+    },
+    {
+      "epoch": 0.6545188272734954,
+      "grad_norm": 0.47902953399381804,
+      "learning_rate": 1.461308419658678e-06,
+      "loss": 0.1468,
+      "step": 2905
+    },
+    {
+      "epoch": 0.6547441349593038,
+      "grad_norm": 0.4592522287454078,
+      "learning_rate": 1.4596142082006448e-06,
+      "loss": 0.1387,
+      "step": 2906
+    },
+    {
+      "epoch": 0.6549694426451123,
+      "grad_norm": 0.5219859102858139,
+      "learning_rate": 1.457920574430062e-06,
+      "loss": 0.1742,
+      "step": 2907
+    },
+    {
+      "epoch": 0.6551947503309207,
+      "grad_norm": 0.485989238913835,
+      "learning_rate": 1.456227519287343e-06,
+      "loss": 0.1563,
+      "step": 2908
+    },
+    {
+      "epoch": 0.6554200580167291,
+      "grad_norm": 0.4718340205076326,
+      "learning_rate": 1.4545350437125755e-06,
+      "loss": 0.1429,
+      "step": 2909
+    },
+    {
+      "epoch": 0.6556453657025375,
+      "grad_norm": 0.4990272336147076,
+      "learning_rate": 1.4528431486455311e-06,
+      "loss": 0.1543,
+      "step": 2910
+    },
+    {
+      "epoch": 0.6558706733883459,
+      "grad_norm": 0.48413560475749645,
+      "learning_rate": 1.451151835025653e-06,
+      "loss": 0.1483,
+      "step": 2911
+    },
+    {
+      "epoch": 0.6560959810741543,
+      "grad_norm": 0.47270868455230225,
+      "learning_rate": 1.4494611037920667e-06,
+      "loss": 0.1396,
+      "step": 2912
+    },
+    {
+      "epoch": 0.6563212887599629,
+      "grad_norm": 0.47527159314782774,
+      "learning_rate": 1.4477709558835724e-06,
+      "loss": 0.1485,
+      "step": 2913
+    },
+    {
+      "epoch": 0.6565465964457713,
+      "grad_norm": 0.4760435028537632,
+      "learning_rate": 1.4460813922386446e-06,
+      "loss": 0.1474,
+      "step": 2914
+    },
+    {
+      "epoch": 0.6567719041315797,
+      "grad_norm": 0.4593543242080812,
+      "learning_rate": 1.4443924137954368e-06,
+      "loss": 0.1367,
+      "step": 2915
+    },
+    {
+      "epoch": 0.6569972118173881,
+      "grad_norm": 0.4850288721265088,
+      "learning_rate": 1.4427040214917742e-06,
+      "loss": 0.1472,
+      "step": 2916
+    },
+    {
+      "epoch": 0.6572225195031965,
+      "grad_norm": 0.4735458603071915,
+      "learning_rate": 1.4410162162651586e-06,
+      "loss": 0.1457,
+      "step": 2917
+    },
+    {
+      "epoch": 0.657447827189005,
+      "grad_norm": 0.4919094654596746,
+      "learning_rate": 1.4393289990527665e-06,
+      "loss": 0.1353,
+      "step": 2918
+    },
+    {
+      "epoch": 0.6576731348748134,
+      "grad_norm": 0.4858859607661872,
+      "learning_rate": 1.4376423707914462e-06,
+      "loss": 0.1494,
+      "step": 2919
+    },
+    {
+      "epoch": 0.6578984425606218,
+      "grad_norm": 0.4690348062287447,
+      "learning_rate": 1.4359563324177176e-06,
+      "loss": 0.138,
+      "step": 2920
+    },
+    {
+      "epoch": 0.6581237502464303,
+      "grad_norm": 0.46664358476670137,
+      "learning_rate": 1.4342708848677774e-06,
+      "loss": 0.1413,
+      "step": 2921
+    },
+    {
+      "epoch": 0.6583490579322387,
+      "grad_norm": 0.46423462010928873,
+      "learning_rate": 1.43258602907749e-06,
+      "loss": 0.1447,
+      "step": 2922
+    },
+    {
+      "epoch": 0.6585743656180472,
+      "grad_norm": 0.4702745421807713,
+      "learning_rate": 1.430901765982395e-06,
+      "loss": 0.1524,
+      "step": 2923
+    },
+    {
+      "epoch": 0.6587996733038556,
+      "grad_norm": 0.48845161374721735,
+      "learning_rate": 1.429218096517699e-06,
+      "loss": 0.1502,
+      "step": 2924
+    },
+    {
+      "epoch": 0.659024980989664,
+      "grad_norm": 0.4176507440012528,
+      "learning_rate": 1.4275350216182824e-06,
+      "loss": 0.1187,
+      "step": 2925
+    },
+    {
+      "epoch": 0.6592502886754724,
+      "grad_norm": 0.4989305833263455,
+      "learning_rate": 1.425852542218692e-06,
+      "loss": 0.1602,
+      "step": 2926
+    },
+    {
+      "epoch": 0.6594755963612808,
+      "grad_norm": 0.47470946640752715,
+      "learning_rate": 1.4241706592531473e-06,
+      "loss": 0.1491,
+      "step": 2927
+    },
+    {
+      "epoch": 0.6597009040470893,
+      "grad_norm": 0.4600229064740794,
+      "learning_rate": 1.4224893736555364e-06,
+      "loss": 0.1499,
+      "step": 2928
+    },
+    {
+      "epoch": 0.6599262117328978,
+      "grad_norm": 0.48995371068382887,
+      "learning_rate": 1.420808686359412e-06,
+      "loss": 0.1559,
+      "step": 2929
+    },
+    {
+      "epoch": 0.6601515194187062,
+      "grad_norm": 0.4815411645705524,
+      "learning_rate": 1.4191285982979992e-06,
+      "loss": 0.1381,
+      "step": 2930
+    },
+    {
+      "epoch": 0.6603768271045146,
+      "grad_norm": 0.47921268405426853,
+      "learning_rate": 1.4174491104041866e-06,
+      "loss": 0.1555,
+      "step": 2931
+    },
+    {
+      "epoch": 0.660602134790323,
+      "grad_norm": 0.48304499533202994,
+      "learning_rate": 1.4157702236105326e-06,
+      "loss": 0.1555,
+      "step": 2932
+    },
+    {
+      "epoch": 0.6608274424761315,
+      "grad_norm": 0.48077288960623543,
+      "learning_rate": 1.414091938849259e-06,
+      "loss": 0.1466,
+      "step": 2933
+    },
+    {
+      "epoch": 0.6610527501619399,
+      "grad_norm": 0.46750831606409243,
+      "learning_rate": 1.412414257052256e-06,
+      "loss": 0.1424,
+      "step": 2934
+    },
+    {
+      "epoch": 0.6612780578477483,
+      "grad_norm": 0.4999024192245585,
+      "learning_rate": 1.410737179151078e-06,
+      "loss": 0.1558,
+      "step": 2935
+    },
+    {
+      "epoch": 0.6615033655335567,
+      "grad_norm": 0.45918538077085974,
+      "learning_rate": 1.4090607060769423e-06,
+      "loss": 0.1368,
+      "step": 2936
+    },
+    {
+      "epoch": 0.6617286732193652,
+      "grad_norm": 0.46695971977298123,
+      "learning_rate": 1.407384838760734e-06,
+      "loss": 0.1516,
+      "step": 2937
+    },
+    {
+      "epoch": 0.6619539809051737,
+      "grad_norm": 0.46204373726139225,
+      "learning_rate": 1.4057095781329983e-06,
+      "loss": 0.1374,
+      "step": 2938
+    },
+    {
+      "epoch": 0.6621792885909821,
+      "grad_norm": 0.4575466919551392,
+      "learning_rate": 1.4040349251239444e-06,
+      "loss": 0.1268,
+      "step": 2939
+    },
+    {
+      "epoch": 0.6624045962767905,
+      "grad_norm": 0.45667917133040004,
+      "learning_rate": 1.402360880663447e-06,
+      "loss": 0.1446,
+      "step": 2940
+    },
+    {
+      "epoch": 0.6626299039625989,
+      "grad_norm": 0.47617239978779796,
+      "learning_rate": 1.4006874456810377e-06,
+      "loss": 0.1508,
+      "step": 2941
+    },
+    {
+      "epoch": 0.6628552116484073,
+      "grad_norm": 0.47224186074775903,
+      "learning_rate": 1.3990146211059141e-06,
+      "loss": 0.1518,
+      "step": 2942
+    },
+    {
+      "epoch": 0.6630805193342157,
+      "grad_norm": 0.4780137911390611,
+      "learning_rate": 1.3973424078669346e-06,
+      "loss": 0.1491,
+      "step": 2943
+    },
+    {
+      "epoch": 0.6633058270200243,
+      "grad_norm": 0.5216227820812853,
+      "learning_rate": 1.3956708068926141e-06,
+      "loss": 0.1639,
+      "step": 2944
+    },
+    {
+      "epoch": 0.6635311347058327,
+      "grad_norm": 0.5137910334803849,
+      "learning_rate": 1.393999819111133e-06,
+      "loss": 0.1488,
+      "step": 2945
+    },
+    {
+      "epoch": 0.6637564423916411,
+      "grad_norm": 0.47392884476866215,
+      "learning_rate": 1.3923294454503263e-06,
+      "loss": 0.1434,
+      "step": 2946
+    },
+    {
+      "epoch": 0.6639817500774495,
+      "grad_norm": 0.46492626782696317,
+      "learning_rate": 1.3906596868376923e-06,
+      "loss": 0.1491,
+      "step": 2947
+    },
+    {
+      "epoch": 0.664207057763258,
+      "grad_norm": 0.44234123269093023,
+      "learning_rate": 1.3889905442003836e-06,
+      "loss": 0.128,
+      "step": 2948
+    },
+    {
+      "epoch": 0.6644323654490664,
+      "grad_norm": 0.48399930599781904,
+      "learning_rate": 1.3873220184652143e-06,
+      "loss": 0.1558,
+      "step": 2949
+    },
+    {
+      "epoch": 0.6646576731348748,
+      "grad_norm": 0.47807200291396135,
+      "learning_rate": 1.3856541105586545e-06,
+      "loss": 0.1526,
+      "step": 2950
+    },
+    {
+      "epoch": 0.6648829808206832,
+      "grad_norm": 0.488387437019316,
+      "learning_rate": 1.3839868214068303e-06,
+      "loss": 0.1495,
+      "step": 2951
+    },
+    {
+      "epoch": 0.6651082885064917,
+      "grad_norm": 0.47430374069411185,
+      "learning_rate": 1.382320151935527e-06,
+      "loss": 0.1384,
+      "step": 2952
+    },
+    {
+      "epoch": 0.6653335961923001,
+      "grad_norm": 0.4753402488559886,
+      "learning_rate": 1.380654103070182e-06,
+      "loss": 0.1364,
+      "step": 2953
+    },
+    {
+      "epoch": 0.6655589038781086,
+      "grad_norm": 0.47795924208162,
+      "learning_rate": 1.3789886757358916e-06,
+      "loss": 0.15,
+      "step": 2954
+    },
+    {
+      "epoch": 0.665784211563917,
+      "grad_norm": 0.49872767832251386,
+      "learning_rate": 1.3773238708574054e-06,
+      "loss": 0.1553,
+      "step": 2955
+    },
+    {
+      "epoch": 0.6660095192497254,
+      "grad_norm": 0.46166735586442514,
+      "learning_rate": 1.375659689359126e-06,
+      "loss": 0.1467,
+      "step": 2956
+    },
+    {
+      "epoch": 0.6662348269355338,
+      "grad_norm": 0.4504446375247205,
+      "learning_rate": 1.3739961321651139e-06,
+      "loss": 0.1413,
+      "step": 2957
+    },
+    {
+      "epoch": 0.6664601346213422,
+      "grad_norm": 0.4831570270140901,
+      "learning_rate": 1.3723332001990774e-06,
+      "loss": 0.1349,
+      "step": 2958
+    },
+    {
+      "epoch": 0.6666854423071507,
+      "grad_norm": 0.43536534788662207,
+      "learning_rate": 1.3706708943843822e-06,
+      "loss": 0.1254,
+      "step": 2959
+    },
+    {
+      "epoch": 0.6669107499929592,
+      "grad_norm": 0.4635897581330502,
+      "learning_rate": 1.369009215644046e-06,
+      "loss": 0.1408,
+      "step": 2960
+    },
+    {
+      "epoch": 0.6671360576787676,
+      "grad_norm": 0.4607848426627536,
+      "learning_rate": 1.3673481649007347e-06,
+      "loss": 0.1413,
+      "step": 2961
+    },
+    {
+      "epoch": 0.667361365364576,
+      "grad_norm": 0.45287392564560036,
+      "learning_rate": 1.36568774307677e-06,
+      "loss": 0.1339,
+      "step": 2962
+    },
+    {
+      "epoch": 0.6675866730503844,
+      "grad_norm": 0.4653937508230895,
+      "learning_rate": 1.36402795109412e-06,
+      "loss": 0.142,
+      "step": 2963
+    },
+    {
+      "epoch": 0.6678119807361929,
+      "grad_norm": 0.4456093510289949,
+      "learning_rate": 1.362368789874407e-06,
+      "loss": 0.133,
+      "step": 2964
+    },
+    {
+      "epoch": 0.6680372884220013,
+      "grad_norm": 0.47497169797980815,
+      "learning_rate": 1.3607102603389016e-06,
+      "loss": 0.1445,
+      "step": 2965
+    },
+    {
+      "epoch": 0.6682625961078097,
+      "grad_norm": 0.4808704744452402,
+      "learning_rate": 1.3590523634085218e-06,
+      "loss": 0.1464,
+      "step": 2966
+    },
+    {
+      "epoch": 0.6684879037936181,
+      "grad_norm": 0.46680295816534506,
+      "learning_rate": 1.3573951000038376e-06,
+      "loss": 0.1351,
+      "step": 2967
+    },
+    {
+      "epoch": 0.6687132114794266,
+      "grad_norm": 0.46521906793161333,
+      "learning_rate": 1.3557384710450644e-06,
+      "loss": 0.159,
+      "step": 2968
+    },
+    {
+      "epoch": 0.6689385191652351,
+      "grad_norm": 0.46029143620950225,
+      "learning_rate": 1.3540824774520678e-06,
+      "loss": 0.1437,
+      "step": 2969
+    },
+    {
+      "epoch": 0.6691638268510435,
+      "grad_norm": 0.4681229343815473,
+      "learning_rate": 1.3524271201443578e-06,
+      "loss": 0.1382,
+      "step": 2970
+    },
+    {
+      "epoch": 0.6693891345368519,
+      "grad_norm": 0.5077166184117899,
+      "learning_rate": 1.3507724000410933e-06,
+      "loss": 0.1567,
+      "step": 2971
+    },
+    {
+      "epoch": 0.6696144422226603,
+      "grad_norm": 0.43495619042875877,
+      "learning_rate": 1.3491183180610807e-06,
+      "loss": 0.1383,
+      "step": 2972
+    },
+    {
+      "epoch": 0.6698397499084687,
+      "grad_norm": 0.45458783126562835,
+      "learning_rate": 1.347464875122766e-06,
+      "loss": 0.134,
+      "step": 2973
+    },
+    {
+      "epoch": 0.6700650575942771,
+      "grad_norm": 0.44400295352887237,
+      "learning_rate": 1.3458120721442464e-06,
+      "loss": 0.1316,
+      "step": 2974
+    },
+    {
+      "epoch": 0.6702903652800856,
+      "grad_norm": 0.45680543296708814,
+      "learning_rate": 1.3441599100432635e-06,
+      "loss": 0.135,
+      "step": 2975
+    },
+    {
+      "epoch": 0.6705156729658941,
+      "grad_norm": 0.46350076737877277,
+      "learning_rate": 1.3425083897371983e-06,
+      "loss": 0.1365,
+      "step": 2976
+    },
+    {
+      "epoch": 0.6707409806517025,
+      "grad_norm": 0.4925752036708255,
+      "learning_rate": 1.3408575121430812e-06,
+      "loss": 0.1456,
+      "step": 2977
+    },
+    {
+      "epoch": 0.6709662883375109,
+      "grad_norm": 0.48149834877938713,
+      "learning_rate": 1.3392072781775806e-06,
+      "loss": 0.1448,
+      "step": 2978
+    },
+    {
+      "epoch": 0.6711915960233193,
+      "grad_norm": 0.44479976304247376,
+      "learning_rate": 1.337557688757012e-06,
+      "loss": 0.1298,
+      "step": 2979
+    },
+    {
+      "epoch": 0.6714169037091278,
+      "grad_norm": 0.49638144928389527,
+      "learning_rate": 1.335908744797329e-06,
+      "loss": 0.1478,
+      "step": 2980
+    },
+    {
+      "epoch": 0.6716422113949362,
+      "grad_norm": 0.496661796314984,
+      "learning_rate": 1.3342604472141296e-06,
+      "loss": 0.154,
+      "step": 2981
+    },
+    {
+      "epoch": 0.6718675190807446,
+      "grad_norm": 0.48523105822115803,
+      "learning_rate": 1.3326127969226535e-06,
+      "loss": 0.1545,
+      "step": 2982
+    },
+    {
+      "epoch": 0.672092826766553,
+      "grad_norm": 0.4706474816730592,
+      "learning_rate": 1.3309657948377768e-06,
+      "loss": 0.1444,
+      "step": 2983
+    },
+    {
+      "epoch": 0.6723181344523615,
+      "grad_norm": 0.46988659672089167,
+      "learning_rate": 1.3293194418740207e-06,
+      "loss": 0.146,
+      "step": 2984
+    },
+    {
+      "epoch": 0.67254344213817,
+      "grad_norm": 0.47750222447061125,
+      "learning_rate": 1.3276737389455416e-06,
+      "loss": 0.1449,
+      "step": 2985
+    },
+    {
+      "epoch": 0.6727687498239784,
+      "grad_norm": 0.4842288786003118,
+      "learning_rate": 1.3260286869661378e-06,
+      "loss": 0.1449,
+      "step": 2986
+    },
+    {
+      "epoch": 0.6729940575097868,
+      "grad_norm": 0.47033341161419007,
+      "learning_rate": 1.3243842868492468e-06,
+      "loss": 0.1364,
+      "step": 2987
+    },
+    {
+      "epoch": 0.6732193651955952,
+      "grad_norm": 0.47443391802769397,
+      "learning_rate": 1.32274053950794e-06,
+      "loss": 0.1436,
+      "step": 2988
+    },
+    {
+      "epoch": 0.6734446728814036,
+      "grad_norm": 0.4731098263985566,
+      "learning_rate": 1.3210974458549318e-06,
+      "loss": 0.1454,
+      "step": 2989
+    },
+    {
+      "epoch": 0.6736699805672121,
+      "grad_norm": 0.46654736287810944,
+      "learning_rate": 1.3194550068025697e-06,
+      "loss": 0.1466,
+      "step": 2990
+    },
+    {
+      "epoch": 0.6738952882530206,
+      "grad_norm": 0.49532064733223036,
+      "learning_rate": 1.3178132232628374e-06,
+      "loss": 0.1478,
+      "step": 2991
+    },
+    {
+      "epoch": 0.674120595938829,
+      "grad_norm": 0.4895887904560521,
+      "learning_rate": 1.3161720961473583e-06,
+      "loss": 0.1505,
+      "step": 2992
+    },
+    {
+      "epoch": 0.6743459036246374,
+      "grad_norm": 0.47395701914670035,
+      "learning_rate": 1.3145316263673874e-06,
+      "loss": 0.1354,
+      "step": 2993
+    },
+    {
+      "epoch": 0.6745712113104458,
+      "grad_norm": 0.49688525999940936,
+      "learning_rate": 1.3128918148338183e-06,
+      "loss": 0.1556,
+      "step": 2994
+    },
+    {
+      "epoch": 0.6747965189962543,
+      "grad_norm": 0.48482779160388284,
+      "learning_rate": 1.3112526624571753e-06,
+      "loss": 0.1534,
+      "step": 2995
+    },
+    {
+      "epoch": 0.6750218266820627,
+      "grad_norm": 0.4970336841742754,
+      "learning_rate": 1.3096141701476189e-06,
+      "loss": 0.1544,
+      "step": 2996
+    },
+    {
+      "epoch": 0.6752471343678711,
+      "grad_norm": 0.49387939090629773,
+      "learning_rate": 1.307976338814945e-06,
+      "loss": 0.1533,
+      "step": 2997
+    },
+    {
+      "epoch": 0.6754724420536795,
+      "grad_norm": 0.4988557664085388,
+      "learning_rate": 1.3063391693685773e-06,
+      "loss": 0.1691,
+      "step": 2998
+    },
+    {
+      "epoch": 0.675697749739488,
+      "grad_norm": 0.45405104965930987,
+      "learning_rate": 1.3047026627175774e-06,
+      "loss": 0.1387,
+      "step": 2999
+    },
+    {
+      "epoch": 0.6759230574252965,
+      "grad_norm": 0.47423882703768094,
+      "learning_rate": 1.3030668197706347e-06,
+      "loss": 0.1417,
+      "step": 3000
+    },
+    {
+      "epoch": 0.6759230574252965,
+      "eval_loss": 0.1473056972026825,
+      "eval_runtime": 56.9201,
+      "eval_samples_per_second": 50.422,
+      "eval_steps_per_second": 6.307,
+      "step": 3000
+    },
+    {
+      "epoch": 0.6761483651111049,
+      "grad_norm": 0.4505313679594971,
+      "learning_rate": 1.3014316414360732e-06,
+      "loss": 0.1344,
+      "step": 3001
+    },
+    {
+      "epoch": 0.6763736727969133,
+      "grad_norm": 0.4559618012667919,
+      "learning_rate": 1.2997971286218448e-06,
+      "loss": 0.1314,
+      "step": 3002
+    },
+    {
+      "epoch": 0.6765989804827217,
+      "grad_norm": 0.47844832165319234,
+      "learning_rate": 1.2981632822355344e-06,
+      "loss": 0.1422,
+      "step": 3003
+    },
+    {
+      "epoch": 0.6768242881685301,
+      "grad_norm": 0.48766877285251875,
+      "learning_rate": 1.2965301031843574e-06,
+      "loss": 0.1468,
+      "step": 3004
+    },
+    {
+      "epoch": 0.6770495958543385,
+      "grad_norm": 0.458622595710425,
+      "learning_rate": 1.294897592375155e-06,
+      "loss": 0.1309,
+      "step": 3005
+    },
+    {
+      "epoch": 0.677274903540147,
+      "grad_norm": 0.46647292917179217,
+      "learning_rate": 1.2932657507144014e-06,
+      "loss": 0.147,
+      "step": 3006
+    },
+    {
+      "epoch": 0.6775002112259555,
+      "grad_norm": 0.4655457149851092,
+      "learning_rate": 1.2916345791081964e-06,
+      "loss": 0.1458,
+      "step": 3007
+    },
+    {
+      "epoch": 0.6777255189117639,
+      "grad_norm": 0.4510539918644756,
+      "learning_rate": 1.2900040784622686e-06,
+      "loss": 0.1437,
+      "step": 3008
+    },
+    {
+      "epoch": 0.6779508265975723,
+      "grad_norm": 0.4524162591560385,
+      "learning_rate": 1.2883742496819751e-06,
+      "loss": 0.1357,
+      "step": 3009
+    },
+    {
+      "epoch": 0.6781761342833807,
+      "grad_norm": 0.4821849125579185,
+      "learning_rate": 1.286745093672298e-06,
+      "loss": 0.1477,
+      "step": 3010
+    },
+    {
+      "epoch": 0.6784014419691892,
+      "grad_norm": 0.49276174937941813,
+      "learning_rate": 1.2851166113378471e-06,
+      "loss": 0.1514,
+      "step": 3011
+    },
+    {
+      "epoch": 0.6786267496549976,
+      "grad_norm": 0.45060096939936306,
+      "learning_rate": 1.2834888035828597e-06,
+      "loss": 0.1368,
+      "step": 3012
+    },
+    {
+      "epoch": 0.678852057340806,
+      "grad_norm": 0.46775710048149644,
+      "learning_rate": 1.2818616713111945e-06,
+      "loss": 0.1487,
+      "step": 3013
+    },
+    {
+      "epoch": 0.6790773650266144,
+      "grad_norm": 0.4624662532836971,
+      "learning_rate": 1.2802352154263392e-06,
+      "loss": 0.1385,
+      "step": 3014
+    },
+    {
+      "epoch": 0.679302672712423,
+      "grad_norm": 0.4537436062939665,
+      "learning_rate": 1.2786094368314023e-06,
+      "loss": 0.1345,
+      "step": 3015
+    },
+    {
+      "epoch": 0.6795279803982314,
+      "grad_norm": 0.508926898966882,
+      "learning_rate": 1.2769843364291202e-06,
+      "loss": 0.1535,
+      "step": 3016
+    },
+    {
+      "epoch": 0.6797532880840398,
+      "grad_norm": 0.4896959769941762,
+      "learning_rate": 1.2753599151218483e-06,
+      "loss": 0.1456,
+      "step": 3017
+    },
+    {
+      "epoch": 0.6799785957698482,
+      "grad_norm": 0.44335947567847883,
+      "learning_rate": 1.2737361738115681e-06,
+      "loss": 0.1262,
+      "step": 3018
+    },
+    {
+      "epoch": 0.6802039034556566,
+      "grad_norm": 0.49082267300728155,
+      "learning_rate": 1.2721131133998837e-06,
+      "loss": 0.1467,
+      "step": 3019
+    },
+    {
+      "epoch": 0.680429211141465,
+      "grad_norm": 0.49221069818725716,
+      "learning_rate": 1.2704907347880185e-06,
+      "loss": 0.1507,
+      "step": 3020
+    },
+    {
+      "epoch": 0.6806545188272735,
+      "grad_norm": 0.48681513514678987,
+      "learning_rate": 1.2688690388768205e-06,
+      "loss": 0.1538,
+      "step": 3021
+    },
+    {
+      "epoch": 0.6808798265130819,
+      "grad_norm": 0.5121121495883596,
+      "learning_rate": 1.2672480265667553e-06,
+      "loss": 0.1539,
+      "step": 3022
+    },
+    {
+      "epoch": 0.6811051341988904,
+      "grad_norm": 0.4926215441022388,
+      "learning_rate": 1.2656276987579118e-06,
+      "loss": 0.1554,
+      "step": 3023
+    },
+    {
+      "epoch": 0.6813304418846988,
+      "grad_norm": 0.48046985165609934,
+      "learning_rate": 1.2640080563499977e-06,
+      "loss": 0.1402,
+      "step": 3024
+    },
+    {
+      "epoch": 0.6815557495705072,
+      "grad_norm": 0.46929362908973327,
+      "learning_rate": 1.2623891002423383e-06,
+      "loss": 0.1438,
+      "step": 3025
+    },
+    {
+      "epoch": 0.6817810572563157,
+      "grad_norm": 0.48803921565185054,
+      "learning_rate": 1.2607708313338818e-06,
+      "loss": 0.1614,
+      "step": 3026
+    },
+    {
+      "epoch": 0.6820063649421241,
+      "grad_norm": 0.5093866941830406,
+      "learning_rate": 1.2591532505231906e-06,
+      "loss": 0.1593,
+      "step": 3027
+    },
+    {
+      "epoch": 0.6822316726279325,
+      "grad_norm": 0.4876402897754427,
+      "learning_rate": 1.2575363587084486e-06,
+      "loss": 0.1444,
+      "step": 3028
+    },
+    {
+      "epoch": 0.6824569803137409,
+      "grad_norm": 0.45939270218193157,
+      "learning_rate": 1.2559201567874554e-06,
+      "loss": 0.1439,
+      "step": 3029
+    },
+    {
+      "epoch": 0.6826822879995493,
+      "grad_norm": 0.4936028113453703,
+      "learning_rate": 1.2543046456576267e-06,
+      "loss": 0.1494,
+      "step": 3030
+    },
+    {
+      "epoch": 0.6829075956853579,
+      "grad_norm": 0.5003307934170121,
+      "learning_rate": 1.252689826215997e-06,
+      "loss": 0.1364,
+      "step": 3031
+    },
+    {
+      "epoch": 0.6831329033711663,
+      "grad_norm": 0.47887429632024237,
+      "learning_rate": 1.2510756993592138e-06,
+      "loss": 0.1411,
+      "step": 3032
+    },
+    {
+      "epoch": 0.6833582110569747,
+      "grad_norm": 0.4563775894109953,
+      "learning_rate": 1.2494622659835421e-06,
+      "loss": 0.1361,
+      "step": 3033
+    },
+    {
+      "epoch": 0.6835835187427831,
+      "grad_norm": 0.4534763623106835,
+      "learning_rate": 1.2478495269848626e-06,
+      "loss": 0.1395,
+      "step": 3034
+    },
+    {
+      "epoch": 0.6838088264285915,
+      "grad_norm": 0.5089995426109665,
+      "learning_rate": 1.246237483258667e-06,
+      "loss": 0.1511,
+      "step": 3035
+    },
+    {
+      "epoch": 0.6840341341144,
+      "grad_norm": 0.47351700691400833,
+      "learning_rate": 1.2446261357000655e-06,
+      "loss": 0.1314,
+      "step": 3036
+    },
+    {
+      "epoch": 0.6842594418002084,
+      "grad_norm": 0.4435009128269627,
+      "learning_rate": 1.243015485203777e-06,
+      "loss": 0.1355,
+      "step": 3037
+    },
+    {
+      "epoch": 0.6844847494860169,
+      "grad_norm": 0.4581160172261627,
+      "learning_rate": 1.2414055326641378e-06,
+      "loss": 0.1481,
+      "step": 3038
+    },
+    {
+      "epoch": 0.6847100571718253,
+      "grad_norm": 0.45155658956140055,
+      "learning_rate": 1.2397962789750923e-06,
+      "loss": 0.1355,
+      "step": 3039
+    },
+    {
+      "epoch": 0.6849353648576337,
+      "grad_norm": 0.4907567897160798,
+      "learning_rate": 1.2381877250302002e-06,
+      "loss": 0.1566,
+      "step": 3040
+    },
+    {
+      "epoch": 0.6851606725434422,
+      "grad_norm": 0.4899167405839578,
+      "learning_rate": 1.236579871722633e-06,
+      "loss": 0.1546,
+      "step": 3041
+    },
+    {
+      "epoch": 0.6853859802292506,
+      "grad_norm": 0.5133340760228167,
+      "learning_rate": 1.2349727199451696e-06,
+      "loss": 0.16,
+      "step": 3042
+    },
+    {
+      "epoch": 0.685611287915059,
+      "grad_norm": 0.45618746757252454,
+      "learning_rate": 1.233366270590202e-06,
+      "loss": 0.1421,
+      "step": 3043
+    },
+    {
+      "epoch": 0.6858365956008674,
+      "grad_norm": 0.5079427568416787,
+      "learning_rate": 1.2317605245497324e-06,
+      "loss": 0.1442,
+      "step": 3044
+    },
+    {
+      "epoch": 0.6860619032866758,
+      "grad_norm": 0.4735908283197607,
+      "learning_rate": 1.2301554827153703e-06,
+      "loss": 0.1444,
+      "step": 3045
+    },
+    {
+      "epoch": 0.6862872109724844,
+      "grad_norm": 0.509591432751062,
+      "learning_rate": 1.2285511459783373e-06,
+      "loss": 0.1619,
+      "step": 3046
+    },
+    {
+      "epoch": 0.6865125186582928,
+      "grad_norm": 0.4331611286622087,
+      "learning_rate": 1.2269475152294601e-06,
+      "loss": 0.1287,
+      "step": 3047
+    },
+    {
+      "epoch": 0.6867378263441012,
+      "grad_norm": 0.4767098179638202,
+      "learning_rate": 1.225344591359177e-06,
+      "loss": 0.1425,
+      "step": 3048
+    },
+    {
+      "epoch": 0.6869631340299096,
+      "grad_norm": 0.4870537149018392,
+      "learning_rate": 1.2237423752575297e-06,
+      "loss": 0.1539,
+      "step": 3049
+    },
+    {
+      "epoch": 0.687188441715718,
+      "grad_norm": 0.4686958612332258,
+      "learning_rate": 1.2221408678141702e-06,
+      "loss": 0.1475,
+      "step": 3050
+    },
+    {
+      "epoch": 0.6874137494015264,
+      "grad_norm": 0.4866315832405193,
+      "learning_rate": 1.220540069918357e-06,
+      "loss": 0.1494,
+      "step": 3051
+    },
+    {
+      "epoch": 0.6876390570873349,
+      "grad_norm": 0.44919615760515846,
+      "learning_rate": 1.2189399824589513e-06,
+      "loss": 0.1344,
+      "step": 3052
+    },
+    {
+      "epoch": 0.6878643647731433,
+      "grad_norm": 0.49054924470874584,
+      "learning_rate": 1.217340606324424e-06,
+      "loss": 0.1619,
+      "step": 3053
+    },
+    {
+      "epoch": 0.6880896724589518,
+      "grad_norm": 0.47122438915104853,
+      "learning_rate": 1.2157419424028473e-06,
+      "loss": 0.1525,
+      "step": 3054
+    },
+    {
+      "epoch": 0.6883149801447602,
+      "grad_norm": 0.47019681817702275,
+      "learning_rate": 1.2141439915819008e-06,
+      "loss": 0.1462,
+      "step": 3055
+    },
+    {
+      "epoch": 0.6885402878305686,
+      "grad_norm": 0.5068235152597098,
+      "learning_rate": 1.2125467547488676e-06,
+      "loss": 0.1657,
+      "step": 3056
+    },
+    {
+      "epoch": 0.6887655955163771,
+      "grad_norm": 0.48158637474706467,
+      "learning_rate": 1.210950232790632e-06,
+      "loss": 0.1539,
+      "step": 3057
+    },
+    {
+      "epoch": 0.6889909032021855,
+      "grad_norm": 0.47591504746128516,
+      "learning_rate": 1.2093544265936848e-06,
+      "loss": 0.1462,
+      "step": 3058
+    },
+    {
+      "epoch": 0.6892162108879939,
+      "grad_norm": 0.48149830718783326,
+      "learning_rate": 1.2077593370441165e-06,
+      "loss": 0.1446,
+      "step": 3059
+    },
+    {
+      "epoch": 0.6894415185738023,
+      "grad_norm": 0.47270146200564156,
+      "learning_rate": 1.206164965027622e-06,
+      "loss": 0.1472,
+      "step": 3060
+    },
+    {
+      "epoch": 0.6896668262596107,
+      "grad_norm": 0.46228603096579945,
+      "learning_rate": 1.204571311429496e-06,
+      "loss": 0.1417,
+      "step": 3061
+    },
+    {
+      "epoch": 0.6898921339454193,
+      "grad_norm": 0.46404709896039853,
+      "learning_rate": 1.2029783771346344e-06,
+      "loss": 0.1498,
+      "step": 3062
+    },
+    {
+      "epoch": 0.6901174416312277,
+      "grad_norm": 0.4831439997979854,
+      "learning_rate": 1.2013861630275353e-06,
+      "loss": 0.1541,
+      "step": 3063
+    },
+    {
+      "epoch": 0.6903427493170361,
+      "grad_norm": 0.4737086389514101,
+      "learning_rate": 1.1997946699922946e-06,
+      "loss": 0.1419,
+      "step": 3064
+    },
+    {
+      "epoch": 0.6905680570028445,
+      "grad_norm": 0.4690439737437469,
+      "learning_rate": 1.1982038989126096e-06,
+      "loss": 0.1466,
+      "step": 3065
+    },
+    {
+      "epoch": 0.6907933646886529,
+      "grad_norm": 0.4594926738760316,
+      "learning_rate": 1.1966138506717776e-06,
+      "loss": 0.1465,
+      "step": 3066
+    },
+    {
+      "epoch": 0.6910186723744614,
+      "grad_norm": 0.4779588042627649,
+      "learning_rate": 1.195024526152691e-06,
+      "loss": 0.1525,
+      "step": 3067
+    },
+    {
+      "epoch": 0.6912439800602698,
+      "grad_norm": 0.4929196065929435,
+      "learning_rate": 1.1934359262378443e-06,
+      "loss": 0.1551,
+      "step": 3068
+    },
+    {
+      "epoch": 0.6914692877460782,
+      "grad_norm": 0.47498771172548415,
+      "learning_rate": 1.1918480518093259e-06,
+      "loss": 0.1418,
+      "step": 3069
+    },
+    {
+      "epoch": 0.6916945954318867,
+      "grad_norm": 0.46425031665508576,
+      "learning_rate": 1.190260903748825e-06,
+      "loss": 0.1239,
+      "step": 3070
+    },
+    {
+      "epoch": 0.6919199031176951,
+      "grad_norm": 0.47318505042928816,
+      "learning_rate": 1.1886744829376243e-06,
+      "loss": 0.1431,
+      "step": 3071
+    },
+    {
+      "epoch": 0.6921452108035036,
+      "grad_norm": 0.4795022202216388,
+      "learning_rate": 1.187088790256605e-06,
+      "loss": 0.1509,
+      "step": 3072
+    },
+    {
+      "epoch": 0.692370518489312,
+      "grad_norm": 0.4894364245128437,
+      "learning_rate": 1.185503826586244e-06,
+      "loss": 0.163,
+      "step": 3073
+    },
+    {
+      "epoch": 0.6925958261751204,
+      "grad_norm": 0.4731165025423434,
+      "learning_rate": 1.1839195928066101e-06,
+      "loss": 0.1475,
+      "step": 3074
+    },
+    {
+      "epoch": 0.6928211338609288,
+      "grad_norm": 0.5098560337269166,
+      "learning_rate": 1.1823360897973723e-06,
+      "loss": 0.166,
+      "step": 3075
+    },
+    {
+      "epoch": 0.6930464415467372,
+      "grad_norm": 0.4565925376100047,
+      "learning_rate": 1.1807533184377882e-06,
+      "loss": 0.1389,
+      "step": 3076
+    },
+    {
+      "epoch": 0.6932717492325456,
+      "grad_norm": 0.4405166749911481,
+      "learning_rate": 1.1791712796067134e-06,
+      "loss": 0.1278,
+      "step": 3077
+    },
+    {
+      "epoch": 0.6934970569183542,
+      "grad_norm": 0.4963434131636802,
+      "learning_rate": 1.1775899741825947e-06,
+      "loss": 0.1474,
+      "step": 3078
+    },
+    {
+      "epoch": 0.6937223646041626,
+      "grad_norm": 0.48453922212336464,
+      "learning_rate": 1.176009403043471e-06,
+      "loss": 0.1501,
+      "step": 3079
+    },
+    {
+      "epoch": 0.693947672289971,
+      "grad_norm": 0.5067028827209051,
+      "learning_rate": 1.1744295670669752e-06,
+      "loss": 0.1496,
+      "step": 3080
+    },
+    {
+      "epoch": 0.6941729799757794,
+      "grad_norm": 0.5177069478578376,
+      "learning_rate": 1.1728504671303326e-06,
+      "loss": 0.1523,
+      "step": 3081
+    },
+    {
+      "epoch": 0.6943982876615878,
+      "grad_norm": 0.4742299577477372,
+      "learning_rate": 1.171272104110356e-06,
+      "loss": 0.1474,
+      "step": 3082
+    },
+    {
+      "epoch": 0.6946235953473963,
+      "grad_norm": 0.4576067845006196,
+      "learning_rate": 1.1696944788834546e-06,
+      "loss": 0.1444,
+      "step": 3083
+    },
+    {
+      "epoch": 0.6948489030332047,
+      "grad_norm": 0.4799393814344733,
+      "learning_rate": 1.168117592325622e-06,
+      "loss": 0.1508,
+      "step": 3084
+    },
+    {
+      "epoch": 0.6950742107190131,
+      "grad_norm": 0.4659138517296435,
+      "learning_rate": 1.1665414453124468e-06,
+      "loss": 0.1438,
+      "step": 3085
+    },
+    {
+      "epoch": 0.6952995184048216,
+      "grad_norm": 0.48305159441188866,
+      "learning_rate": 1.1649660387191027e-06,
+      "loss": 0.1427,
+      "step": 3086
+    },
+    {
+      "epoch": 0.69552482609063,
+      "grad_norm": 0.4772218284362722,
+      "learning_rate": 1.1633913734203552e-06,
+      "loss": 0.149,
+      "step": 3087
+    },
+    {
+      "epoch": 0.6957501337764385,
+      "grad_norm": 0.45069555823830265,
+      "learning_rate": 1.1618174502905586e-06,
+      "loss": 0.1402,
+      "step": 3088
+    },
+    {
+      "epoch": 0.6959754414622469,
+      "grad_norm": 0.5075563908277877,
+      "learning_rate": 1.1602442702036513e-06,
+      "loss": 0.1573,
+      "step": 3089
+    },
+    {
+      "epoch": 0.6962007491480553,
+      "grad_norm": 0.49676498026711446,
+      "learning_rate": 1.1586718340331634e-06,
+      "loss": 0.1502,
+      "step": 3090
+    },
+    {
+      "epoch": 0.6964260568338637,
+      "grad_norm": 0.4874275007939701,
+      "learning_rate": 1.1571001426522088e-06,
+      "loss": 0.1445,
+      "step": 3091
+    },
+    {
+      "epoch": 0.6966513645196721,
+      "grad_norm": 0.48490688683211536,
+      "learning_rate": 1.1555291969334907e-06,
+      "loss": 0.1508,
+      "step": 3092
+    },
+    {
+      "epoch": 0.6968766722054807,
+      "grad_norm": 0.4492305244596637,
+      "learning_rate": 1.1539589977492946e-06,
+      "loss": 0.1227,
+      "step": 3093
+    },
+    {
+      "epoch": 0.6971019798912891,
+      "grad_norm": 0.5051266177903154,
+      "learning_rate": 1.1523895459714948e-06,
+      "loss": 0.1545,
+      "step": 3094
+    },
+    {
+      "epoch": 0.6973272875770975,
+      "grad_norm": 0.4926487663638566,
+      "learning_rate": 1.1508208424715511e-06,
+      "loss": 0.153,
+      "step": 3095
+    },
+    {
+      "epoch": 0.6975525952629059,
+      "grad_norm": 0.4848838310155643,
+      "learning_rate": 1.1492528881205027e-06,
+      "loss": 0.1451,
+      "step": 3096
+    },
+    {
+      "epoch": 0.6977779029487143,
+      "grad_norm": 0.46075829468760326,
+      "learning_rate": 1.1476856837889774e-06,
+      "loss": 0.1488,
+      "step": 3097
+    },
+    {
+      "epoch": 0.6980032106345228,
+      "grad_norm": 0.49275519224627967,
+      "learning_rate": 1.146119230347187e-06,
+      "loss": 0.15,
+      "step": 3098
+    },
+    {
+      "epoch": 0.6982285183203312,
+      "grad_norm": 0.4629973565314605,
+      "learning_rate": 1.1445535286649223e-06,
+      "loss": 0.1393,
+      "step": 3099
+    },
+    {
+      "epoch": 0.6984538260061396,
+      "grad_norm": 0.45565503946869035,
+      "learning_rate": 1.142988579611561e-06,
+      "loss": 0.1418,
+      "step": 3100
+    },
+    {
+      "epoch": 0.6986791336919481,
+      "grad_norm": 0.4760860016759647,
+      "learning_rate": 1.1414243840560595e-06,
+      "loss": 0.1445,
+      "step": 3101
+    },
+    {
+      "epoch": 0.6989044413777565,
+      "grad_norm": 0.46767206653780474,
+      "learning_rate": 1.1398609428669582e-06,
+      "loss": 0.1498,
+      "step": 3102
+    },
+    {
+      "epoch": 0.699129749063565,
+      "grad_norm": 0.4712615428373321,
+      "learning_rate": 1.1382982569123781e-06,
+      "loss": 0.1312,
+      "step": 3103
+    },
+    {
+      "epoch": 0.6993550567493734,
+      "grad_norm": 0.4831477550551603,
+      "learning_rate": 1.136736327060019e-06,
+      "loss": 0.1421,
+      "step": 3104
+    },
+    {
+      "epoch": 0.6995803644351818,
+      "grad_norm": 0.49741492983541774,
+      "learning_rate": 1.1351751541771644e-06,
+      "loss": 0.1535,
+      "step": 3105
+    },
+    {
+      "epoch": 0.6998056721209902,
+      "grad_norm": 0.48827851063729416,
+      "learning_rate": 1.133614739130673e-06,
+      "loss": 0.1436,
+      "step": 3106
+    },
+    {
+      "epoch": 0.7000309798067986,
+      "grad_norm": 0.49902832291234195,
+      "learning_rate": 1.1320550827869875e-06,
+      "loss": 0.1606,
+      "step": 3107
+    },
+    {
+      "epoch": 0.700256287492607,
+      "grad_norm": 0.5116087479087541,
+      "learning_rate": 1.1304961860121246e-06,
+      "loss": 0.1691,
+      "step": 3108
+    },
+    {
+      "epoch": 0.7004815951784156,
+      "grad_norm": 0.510097472402678,
+      "learning_rate": 1.128938049671683e-06,
+      "loss": 0.1616,
+      "step": 3109
+    },
+    {
+      "epoch": 0.700706902864224,
+      "grad_norm": 0.4833713433695349,
+      "learning_rate": 1.127380674630838e-06,
+      "loss": 0.152,
+      "step": 3110
+    },
+    {
+      "epoch": 0.7009322105500324,
+      "grad_norm": 0.47423236577069905,
+      "learning_rate": 1.1258240617543407e-06,
+      "loss": 0.1496,
+      "step": 3111
+    },
+    {
+      "epoch": 0.7011575182358408,
+      "grad_norm": 0.4613882318970388,
+      "learning_rate": 1.1242682119065217e-06,
+      "loss": 0.14,
+      "step": 3112
+    },
+    {
+      "epoch": 0.7013828259216492,
+      "grad_norm": 0.5157137088816274,
+      "learning_rate": 1.1227131259512857e-06,
+      "loss": 0.1586,
+      "step": 3113
+    },
+    {
+      "epoch": 0.7016081336074577,
+      "grad_norm": 0.49516447557609644,
+      "learning_rate": 1.121158804752113e-06,
+      "loss": 0.1527,
+      "step": 3114
+    },
+    {
+      "epoch": 0.7018334412932661,
+      "grad_norm": 0.4952901654808642,
+      "learning_rate": 1.119605249172062e-06,
+      "loss": 0.1513,
+      "step": 3115
+    },
+    {
+      "epoch": 0.7020587489790745,
+      "grad_norm": 0.4724530964207107,
+      "learning_rate": 1.1180524600737624e-06,
+      "loss": 0.1443,
+      "step": 3116
+    },
+    {
+      "epoch": 0.702284056664883,
+      "grad_norm": 0.4810529386464414,
+      "learning_rate": 1.1165004383194218e-06,
+      "loss": 0.1541,
+      "step": 3117
+    },
+    {
+      "epoch": 0.7025093643506914,
+      "grad_norm": 0.49948631908338825,
+      "learning_rate": 1.1149491847708186e-06,
+      "loss": 0.1446,
+      "step": 3118
+    },
+    {
+      "epoch": 0.7027346720364999,
+      "grad_norm": 0.46435650648448656,
+      "learning_rate": 1.1133987002893062e-06,
+      "loss": 0.1402,
+      "step": 3119
+    },
+    {
+      "epoch": 0.7029599797223083,
+      "grad_norm": 0.4748936522784897,
+      "learning_rate": 1.1118489857358129e-06,
+      "loss": 0.1434,
+      "step": 3120
+    },
+    {
+      "epoch": 0.7031852874081167,
+      "grad_norm": 0.4742386565536852,
+      "learning_rate": 1.1103000419708347e-06,
+      "loss": 0.1406,
+      "step": 3121
+    },
+    {
+      "epoch": 0.7034105950939251,
+      "grad_norm": 0.47132136003416986,
+      "learning_rate": 1.1087518698544444e-06,
+      "loss": 0.145,
+      "step": 3122
+    },
+    {
+      "epoch": 0.7036359027797335,
+      "grad_norm": 0.48013507541614425,
+      "learning_rate": 1.1072044702462825e-06,
+      "loss": 0.1581,
+      "step": 3123
+    },
+    {
+      "epoch": 0.703861210465542,
+      "grad_norm": 0.49721536280253587,
+      "learning_rate": 1.1056578440055631e-06,
+      "loss": 0.153,
+      "step": 3124
+    },
+    {
+      "epoch": 0.7040865181513505,
+      "grad_norm": 0.49307643608270285,
+      "learning_rate": 1.1041119919910715e-06,
+      "loss": 0.148,
+      "step": 3125
+    },
+    {
+      "epoch": 0.7043118258371589,
+      "grad_norm": 0.47170175416938803,
+      "learning_rate": 1.1025669150611594e-06,
+      "loss": 0.1481,
+      "step": 3126
+    },
+    {
+      "epoch": 0.7045371335229673,
+      "grad_norm": 0.48379362267533554,
+      "learning_rate": 1.101022614073752e-06,
+      "loss": 0.1413,
+      "step": 3127
+    },
+    {
+      "epoch": 0.7047624412087757,
+      "grad_norm": 0.4795946087501419,
+      "learning_rate": 1.0994790898863409e-06,
+      "loss": 0.1521,
+      "step": 3128
+    },
+    {
+      "epoch": 0.7049877488945842,
+      "grad_norm": 0.4634515948168377,
+      "learning_rate": 1.0979363433559892e-06,
+      "loss": 0.1386,
+      "step": 3129
+    },
+    {
+      "epoch": 0.7052130565803926,
+      "grad_norm": 0.4671318265320663,
+      "learning_rate": 1.0963943753393252e-06,
+      "loss": 0.1414,
+      "step": 3130
+    },
+    {
+      "epoch": 0.705438364266201,
+      "grad_norm": 0.46605061932981395,
+      "learning_rate": 1.094853186692546e-06,
+      "loss": 0.1313,
+      "step": 3131
+    },
+    {
+      "epoch": 0.7056636719520094,
+      "grad_norm": 0.445158905701931,
+      "learning_rate": 1.0933127782714175e-06,
+      "loss": 0.1295,
+      "step": 3132
+    },
+    {
+      "epoch": 0.7058889796378179,
+      "grad_norm": 0.46717957236013813,
+      "learning_rate": 1.0917731509312696e-06,
+      "loss": 0.1419,
+      "step": 3133
+    },
+    {
+      "epoch": 0.7061142873236264,
+      "grad_norm": 0.48689326140673095,
+      "learning_rate": 1.0902343055270006e-06,
+      "loss": 0.1519,
+      "step": 3134
+    },
+    {
+      "epoch": 0.7063395950094348,
+      "grad_norm": 0.46153458245075457,
+      "learning_rate": 1.0886962429130754e-06,
+      "loss": 0.1368,
+      "step": 3135
+    },
+    {
+      "epoch": 0.7065649026952432,
+      "grad_norm": 0.4859316993075991,
+      "learning_rate": 1.0871589639435204e-06,
+      "loss": 0.1458,
+      "step": 3136
+    },
+    {
+      "epoch": 0.7067902103810516,
+      "grad_norm": 0.481466474298156,
+      "learning_rate": 1.0856224694719313e-06,
+      "loss": 0.1458,
+      "step": 3137
+    },
+    {
+      "epoch": 0.70701551806686,
+      "grad_norm": 0.5198663823698088,
+      "learning_rate": 1.0840867603514648e-06,
+      "loss": 0.1662,
+      "step": 3138
+    },
+    {
+      "epoch": 0.7072408257526684,
+      "grad_norm": 0.5043701860625105,
+      "learning_rate": 1.0825518374348442e-06,
+      "loss": 0.1592,
+      "step": 3139
+    },
+    {
+      "epoch": 0.707466133438477,
+      "grad_norm": 0.4683284619656246,
+      "learning_rate": 1.0810177015743536e-06,
+      "loss": 0.141,
+      "step": 3140
+    },
+    {
+      "epoch": 0.7076914411242854,
+      "grad_norm": 0.4772402808000686,
+      "learning_rate": 1.079484353621842e-06,
+      "loss": 0.1471,
+      "step": 3141
+    },
+    {
+      "epoch": 0.7079167488100938,
+      "grad_norm": 0.4826855257837391,
+      "learning_rate": 1.0779517944287216e-06,
+      "loss": 0.1441,
+      "step": 3142
+    },
+    {
+      "epoch": 0.7081420564959022,
+      "grad_norm": 0.4435103670159692,
+      "learning_rate": 1.0764200248459633e-06,
+      "loss": 0.1283,
+      "step": 3143
+    },
+    {
+      "epoch": 0.7083673641817106,
+      "grad_norm": 0.4574956644323345,
+      "learning_rate": 1.0748890457241037e-06,
+      "loss": 0.1398,
+      "step": 3144
+    },
+    {
+      "epoch": 0.7085926718675191,
+      "grad_norm": 0.5063357865724047,
+      "learning_rate": 1.0733588579132365e-06,
+      "loss": 0.1504,
+      "step": 3145
+    },
+    {
+      "epoch": 0.7088179795533275,
+      "grad_norm": 0.47469368780329463,
+      "learning_rate": 1.0718294622630188e-06,
+      "loss": 0.1462,
+      "step": 3146
+    },
+    {
+      "epoch": 0.7090432872391359,
+      "grad_norm": 0.4680882762833536,
+      "learning_rate": 1.0703008596226692e-06,
+      "loss": 0.1352,
+      "step": 3147
+    },
+    {
+      "epoch": 0.7092685949249444,
+      "grad_norm": 0.47703364881096866,
+      "learning_rate": 1.0687730508409594e-06,
+      "loss": 0.157,
+      "step": 3148
+    },
+    {
+      "epoch": 0.7094939026107528,
+      "grad_norm": 0.4786367298568394,
+      "learning_rate": 1.0672460367662271e-06,
+      "loss": 0.1422,
+      "step": 3149
+    },
+    {
+      "epoch": 0.7097192102965613,
+      "grad_norm": 0.4744643950982306,
+      "learning_rate": 1.065719818246367e-06,
+      "loss": 0.146,
+      "step": 3150
+    },
+    {
+      "epoch": 0.7099445179823697,
+      "grad_norm": 0.5116813633903049,
+      "learning_rate": 1.0641943961288298e-06,
+      "loss": 0.1512,
+      "step": 3151
+    },
+    {
+      "epoch": 0.7101698256681781,
+      "grad_norm": 0.48955853146110767,
+      "learning_rate": 1.062669771260627e-06,
+      "loss": 0.1552,
+      "step": 3152
+    },
+    {
+      "epoch": 0.7103951333539865,
+      "grad_norm": 0.4676410023274031,
+      "learning_rate": 1.0611459444883243e-06,
+      "loss": 0.1434,
+      "step": 3153
+    },
+    {
+      "epoch": 0.7106204410397949,
+      "grad_norm": 0.47628413236442235,
+      "learning_rate": 1.0596229166580477e-06,
+      "loss": 0.149,
+      "step": 3154
+    },
+    {
+      "epoch": 0.7108457487256034,
+      "grad_norm": 0.4838018330638107,
+      "learning_rate": 1.0581006886154758e-06,
+      "loss": 0.1413,
+      "step": 3155
+    },
+    {
+      "epoch": 0.7110710564114119,
+      "grad_norm": 0.4772983055827193,
+      "learning_rate": 1.0565792612058462e-06,
+      "loss": 0.148,
+      "step": 3156
+    },
+    {
+      "epoch": 0.7112963640972203,
+      "grad_norm": 0.4804248864966532,
+      "learning_rate": 1.0550586352739519e-06,
+      "loss": 0.147,
+      "step": 3157
+    },
+    {
+      "epoch": 0.7115216717830287,
+      "grad_norm": 0.4494515513060731,
+      "learning_rate": 1.0535388116641376e-06,
+      "loss": 0.1391,
+      "step": 3158
+    },
+    {
+      "epoch": 0.7117469794688371,
+      "grad_norm": 0.4538366196412275,
+      "learning_rate": 1.0520197912203067e-06,
+      "loss": 0.1226,
+      "step": 3159
+    },
+    {
+      "epoch": 0.7119722871546456,
+      "grad_norm": 0.4981133094727563,
+      "learning_rate": 1.050501574785913e-06,
+      "loss": 0.1569,
+      "step": 3160
+    },
+    {
+      "epoch": 0.712197594840454,
+      "grad_norm": 0.4939787669612392,
+      "learning_rate": 1.048984163203967e-06,
+      "loss": 0.1529,
+      "step": 3161
+    },
+    {
+      "epoch": 0.7124229025262624,
+      "grad_norm": 0.480626233851763,
+      "learning_rate": 1.0474675573170293e-06,
+      "loss": 0.1436,
+      "step": 3162
+    },
+    {
+      "epoch": 0.7126482102120708,
+      "grad_norm": 0.47851850089311315,
+      "learning_rate": 1.045951757967215e-06,
+      "loss": 0.1511,
+      "step": 3163
+    },
+    {
+      "epoch": 0.7128735178978793,
+      "grad_norm": 0.4821007601563941,
+      "learning_rate": 1.0444367659961927e-06,
+      "loss": 0.1517,
+      "step": 3164
+    },
+    {
+      "epoch": 0.7130988255836878,
+      "grad_norm": 0.4508947909552097,
+      "learning_rate": 1.0429225822451792e-06,
+      "loss": 0.1427,
+      "step": 3165
+    },
+    {
+      "epoch": 0.7133241332694962,
+      "grad_norm": 0.4908658910207534,
+      "learning_rate": 1.041409207554944e-06,
+      "loss": 0.1469,
+      "step": 3166
+    },
+    {
+      "epoch": 0.7135494409553046,
+      "grad_norm": 0.48281352061819466,
+      "learning_rate": 1.0398966427658091e-06,
+      "loss": 0.1463,
+      "step": 3167
+    },
+    {
+      "epoch": 0.713774748641113,
+      "grad_norm": 0.5024966516256968,
+      "learning_rate": 1.0383848887176437e-06,
+      "loss": 0.155,
+      "step": 3168
+    },
+    {
+      "epoch": 0.7140000563269214,
+      "grad_norm": 0.46731125814676394,
+      "learning_rate": 1.0368739462498704e-06,
+      "loss": 0.1354,
+      "step": 3169
+    },
+    {
+      "epoch": 0.7142253640127298,
+      "grad_norm": 0.48861777845760407,
+      "learning_rate": 1.035363816201457e-06,
+      "loss": 0.1386,
+      "step": 3170
+    },
+    {
+      "epoch": 0.7144506716985383,
+      "grad_norm": 0.4798593003406232,
+      "learning_rate": 1.033854499410924e-06,
+      "loss": 0.141,
+      "step": 3171
+    },
+    {
+      "epoch": 0.7146759793843468,
+      "grad_norm": 0.4951005837529604,
+      "learning_rate": 1.032345996716339e-06,
+      "loss": 0.1376,
+      "step": 3172
+    },
+    {
+      "epoch": 0.7149012870701552,
+      "grad_norm": 0.4649460338111168,
+      "learning_rate": 1.030838308955316e-06,
+      "loss": 0.1302,
+      "step": 3173
+    },
+    {
+      "epoch": 0.7151265947559636,
+      "grad_norm": 0.5208561892477044,
+      "learning_rate": 1.0293314369650193e-06,
+      "loss": 0.1605,
+      "step": 3174
+    },
+    {
+      "epoch": 0.715351902441772,
+      "grad_norm": 0.4446055841305565,
+      "learning_rate": 1.027825381582157e-06,
+      "loss": 0.1237,
+      "step": 3175
+    },
+    {
+      "epoch": 0.7155772101275805,
+      "grad_norm": 0.4780739117334455,
+      "learning_rate": 1.0263201436429873e-06,
+      "loss": 0.1475,
+      "step": 3176
+    },
+    {
+      "epoch": 0.7158025178133889,
+      "grad_norm": 0.45659031396739413,
+      "learning_rate": 1.0248157239833111e-06,
+      "loss": 0.1351,
+      "step": 3177
+    },
+    {
+      "epoch": 0.7160278254991973,
+      "grad_norm": 0.49335546136403297,
+      "learning_rate": 1.0233121234384777e-06,
+      "loss": 0.1482,
+      "step": 3178
+    },
+    {
+      "epoch": 0.7162531331850057,
+      "grad_norm": 0.48359513736268783,
+      "learning_rate": 1.0218093428433807e-06,
+      "loss": 0.1496,
+      "step": 3179
+    },
+    {
+      "epoch": 0.7164784408708142,
+      "grad_norm": 0.47609665715625343,
+      "learning_rate": 1.0203073830324566e-06,
+      "loss": 0.1498,
+      "step": 3180
+    },
+    {
+      "epoch": 0.7167037485566227,
+      "grad_norm": 0.4929326238522347,
+      "learning_rate": 1.0188062448396897e-06,
+      "loss": 0.1454,
+      "step": 3181
+    },
+    {
+      "epoch": 0.7169290562424311,
+      "grad_norm": 0.502710848277265,
+      "learning_rate": 1.0173059290986048e-06,
+      "loss": 0.1369,
+      "step": 3182
+    },
+    {
+      "epoch": 0.7171543639282395,
+      "grad_norm": 0.5268512619847401,
+      "learning_rate": 1.015806436642271e-06,
+      "loss": 0.1569,
+      "step": 3183
+    },
+    {
+      "epoch": 0.7173796716140479,
+      "grad_norm": 0.4880551435141198,
+      "learning_rate": 1.0143077683033017e-06,
+      "loss": 0.1442,
+      "step": 3184
+    },
+    {
+      "epoch": 0.7176049792998563,
+      "grad_norm": 0.5003825145945356,
+      "learning_rate": 1.0128099249138502e-06,
+      "loss": 0.1599,
+      "step": 3185
+    },
+    {
+      "epoch": 0.7178302869856648,
+      "grad_norm": 0.4862755540205847,
+      "learning_rate": 1.0113129073056149e-06,
+      "loss": 0.1574,
+      "step": 3186
+    },
+    {
+      "epoch": 0.7180555946714733,
+      "grad_norm": 0.4802506303500911,
+      "learning_rate": 1.0098167163098319e-06,
+      "loss": 0.1495,
+      "step": 3187
+    },
+    {
+      "epoch": 0.7182809023572817,
+      "grad_norm": 0.499625726077395,
+      "learning_rate": 1.008321352757281e-06,
+      "loss": 0.1525,
+      "step": 3188
+    },
+    {
+      "epoch": 0.7185062100430901,
+      "grad_norm": 0.48435055671852,
+      "learning_rate": 1.0068268174782833e-06,
+      "loss": 0.1593,
+      "step": 3189
+    },
+    {
+      "epoch": 0.7187315177288985,
+      "grad_norm": 0.4921873230317497,
+      "learning_rate": 1.0053331113026962e-06,
+      "loss": 0.1482,
+      "step": 3190
+    },
+    {
+      "epoch": 0.718956825414707,
+      "grad_norm": 0.4612564545959869,
+      "learning_rate": 1.0038402350599214e-06,
+      "loss": 0.1421,
+      "step": 3191
+    },
+    {
+      "epoch": 0.7191821331005154,
+      "grad_norm": 0.5063759140786678,
+      "learning_rate": 1.002348189578895e-06,
+      "loss": 0.1563,
+      "step": 3192
+    },
+    {
+      "epoch": 0.7194074407863238,
+      "grad_norm": 0.46213153113966987,
+      "learning_rate": 1.0008569756880956e-06,
+      "loss": 0.1378,
+      "step": 3193
+    },
+    {
+      "epoch": 0.7196327484721322,
+      "grad_norm": 0.4691935929001728,
+      "learning_rate": 9.993665942155395e-07,
+      "loss": 0.1405,
+      "step": 3194
+    },
+    {
+      "epoch": 0.7198580561579407,
+      "grad_norm": 0.47725314891400694,
+      "learning_rate": 9.978770459887778e-07,
+      "loss": 0.1449,
+      "step": 3195
+    },
+    {
+      "epoch": 0.7200833638437492,
+      "grad_norm": 0.4993885829727644,
+      "learning_rate": 9.963883318349039e-07,
+      "loss": 0.1589,
+      "step": 3196
+    },
+    {
+      "epoch": 0.7203086715295576,
+      "grad_norm": 0.47496255926004993,
+      "learning_rate": 9.949004525805423e-07,
+      "loss": 0.1372,
+      "step": 3197
+    },
+    {
+      "epoch": 0.720533979215366,
+      "grad_norm": 0.42669134354141913,
+      "learning_rate": 9.934134090518593e-07,
+      "loss": 0.1199,
+      "step": 3198
+    },
+    {
+      "epoch": 0.7207592869011744,
+      "grad_norm": 0.5020968589992666,
+      "learning_rate": 9.919272020745529e-07,
+      "loss": 0.1463,
+      "step": 3199
+    },
+    {
+      "epoch": 0.7209845945869828,
+      "grad_norm": 0.4591744042346271,
+      "learning_rate": 9.904418324738605e-07,
+      "loss": 0.1406,
+      "step": 3200
+    },
+    {
+      "epoch": 0.7212099022727912,
+      "grad_norm": 0.47887507892203496,
+      "learning_rate": 9.889573010745507e-07,
+      "loss": 0.1482,
+      "step": 3201
+    },
+    {
+      "epoch": 0.7214352099585997,
+      "grad_norm": 0.4844452159745273,
+      "learning_rate": 9.874736087009285e-07,
+      "loss": 0.1431,
+      "step": 3202
+    },
+    {
+      "epoch": 0.7216605176444082,
+      "grad_norm": 0.4680725094888077,
+      "learning_rate": 9.859907561768335e-07,
+      "loss": 0.1436,
+      "step": 3203
+    },
+    {
+      "epoch": 0.7218858253302166,
+      "grad_norm": 0.4610780009513288,
+      "learning_rate": 9.84508744325639e-07,
+      "loss": 0.136,
+      "step": 3204
+    },
+    {
+      "epoch": 0.722111133016025,
+      "grad_norm": 0.45728936356077726,
+      "learning_rate": 9.830275739702497e-07,
+      "loss": 0.1354,
+      "step": 3205
+    },
+    {
+      "epoch": 0.7223364407018334,
+      "grad_norm": 0.4844421080714476,
+      "learning_rate": 9.815472459331061e-07,
+      "loss": 0.1456,
+      "step": 3206
+    },
+    {
+      "epoch": 0.7225617483876419,
+      "grad_norm": 0.4486188705271111,
+      "learning_rate": 9.800677610361768e-07,
+      "loss": 0.1382,
+      "step": 3207
+    },
+    {
+      "epoch": 0.7227870560734503,
+      "grad_norm": 0.461905229367304,
+      "learning_rate": 9.785891201009667e-07,
+      "loss": 0.1348,
+      "step": 3208
+    },
+    {
+      "epoch": 0.7230123637592587,
+      "grad_norm": 0.4851114349292712,
+      "learning_rate": 9.771113239485084e-07,
+      "loss": 0.154,
+      "step": 3209
+    },
+    {
+      "epoch": 0.7232376714450671,
+      "grad_norm": 0.5133236201683371,
+      "learning_rate": 9.756343733993679e-07,
+      "loss": 0.1604,
+      "step": 3210
+    },
+    {
+      "epoch": 0.7234629791308756,
+      "grad_norm": 0.48957254607104733,
+      "learning_rate": 9.741582692736412e-07,
+      "loss": 0.1438,
+      "step": 3211
+    },
+    {
+      "epoch": 0.7236882868166841,
+      "grad_norm": 0.5149125999254558,
+      "learning_rate": 9.726830123909527e-07,
+      "loss": 0.1542,
+      "step": 3212
+    },
+    {
+      "epoch": 0.7239135945024925,
+      "grad_norm": 0.4900800650716252,
+      "learning_rate": 9.71208603570459e-07,
+      "loss": 0.1476,
+      "step": 3213
+    },
+    {
+      "epoch": 0.7241389021883009,
+      "grad_norm": 0.4618249551389436,
+      "learning_rate": 9.697350436308428e-07,
+      "loss": 0.1448,
+      "step": 3214
+    },
+    {
+      "epoch": 0.7243642098741093,
+      "grad_norm": 0.47297321499334205,
+      "learning_rate": 9.68262333390318e-07,
+      "loss": 0.138,
+      "step": 3215
+    },
+    {
+      "epoch": 0.7245895175599177,
+      "grad_norm": 0.5044794941673408,
+      "learning_rate": 9.667904736666258e-07,
+      "loss": 0.1594,
+      "step": 3216
+    },
+    {
+      "epoch": 0.7248148252457262,
+      "grad_norm": 0.4703026023000923,
+      "learning_rate": 9.653194652770343e-07,
+      "loss": 0.1399,
+      "step": 3217
+    },
+    {
+      "epoch": 0.7250401329315346,
+      "grad_norm": 0.45240034715113475,
+      "learning_rate": 9.638493090383408e-07,
+      "loss": 0.1356,
+      "step": 3218
+    },
+    {
+      "epoch": 0.7252654406173431,
+      "grad_norm": 0.492964125740421,
+      "learning_rate": 9.623800057668675e-07,
+      "loss": 0.1478,
+      "step": 3219
+    },
+    {
+      "epoch": 0.7254907483031515,
+      "grad_norm": 0.4937862053665514,
+      "learning_rate": 9.60911556278463e-07,
+      "loss": 0.1457,
+      "step": 3220
+    },
+    {
+      "epoch": 0.7257160559889599,
+      "grad_norm": 0.4745847894532621,
+      "learning_rate": 9.594439613885044e-07,
+      "loss": 0.1424,
+      "step": 3221
+    },
+    {
+      "epoch": 0.7259413636747684,
+      "grad_norm": 0.47954034109239907,
+      "learning_rate": 9.579772219118899e-07,
+      "loss": 0.1316,
+      "step": 3222
+    },
+    {
+      "epoch": 0.7261666713605768,
+      "grad_norm": 0.5004712882029183,
+      "learning_rate": 9.565113386630482e-07,
+      "loss": 0.1442,
+      "step": 3223
+    },
+    {
+      "epoch": 0.7263919790463852,
+      "grad_norm": 0.5055191565679837,
+      "learning_rate": 9.550463124559267e-07,
+      "loss": 0.1499,
+      "step": 3224
+    },
+    {
+      "epoch": 0.7266172867321936,
+      "grad_norm": 0.4833525032509096,
+      "learning_rate": 9.535821441040017e-07,
+      "loss": 0.1517,
+      "step": 3225
+    },
+    {
+      "epoch": 0.726842594418002,
+      "grad_norm": 0.5025779445176393,
+      "learning_rate": 9.521188344202717e-07,
+      "loss": 0.1567,
+      "step": 3226
+    },
+    {
+      "epoch": 0.7270679021038106,
+      "grad_norm": 0.47225884260255974,
+      "learning_rate": 9.506563842172565e-07,
+      "loss": 0.1418,
+      "step": 3227
+    },
+    {
+      "epoch": 0.727293209789619,
+      "grad_norm": 0.4845795579559558,
+      "learning_rate": 9.491947943070015e-07,
+      "loss": 0.149,
+      "step": 3228
+    },
+    {
+      "epoch": 0.7275185174754274,
+      "grad_norm": 0.47410389383751206,
+      "learning_rate": 9.477340655010717e-07,
+      "loss": 0.1405,
+      "step": 3229
+    },
+    {
+      "epoch": 0.7277438251612358,
+      "grad_norm": 0.48915856141275144,
+      "learning_rate": 9.462741986105573e-07,
+      "loss": 0.1544,
+      "step": 3230
+    },
+    {
+      "epoch": 0.7279691328470442,
+      "grad_norm": 0.4894249277430562,
+      "learning_rate": 9.448151944460657e-07,
+      "loss": 0.1431,
+      "step": 3231
+    },
+    {
+      "epoch": 0.7281944405328526,
+      "grad_norm": 0.45973320661181566,
+      "learning_rate": 9.433570538177289e-07,
+      "loss": 0.1311,
+      "step": 3232
+    },
+    {
+      "epoch": 0.7284197482186611,
+      "grad_norm": 0.4927844223585169,
+      "learning_rate": 9.418997775351985e-07,
+      "loss": 0.1597,
+      "step": 3233
+    },
+    {
+      "epoch": 0.7286450559044696,
+      "grad_norm": 0.4916102224236364,
+      "learning_rate": 9.404433664076442e-07,
+      "loss": 0.148,
+      "step": 3234
+    },
+    {
+      "epoch": 0.728870363590278,
+      "grad_norm": 0.4896074070395624,
+      "learning_rate": 9.389878212437586e-07,
+      "loss": 0.135,
+      "step": 3235
+    },
+    {
+      "epoch": 0.7290956712760864,
+      "grad_norm": 0.49522223597253284,
+      "learning_rate": 9.375331428517506e-07,
+      "loss": 0.1433,
+      "step": 3236
+    },
+    {
+      "epoch": 0.7293209789618949,
+      "grad_norm": 0.49628918046133014,
+      "learning_rate": 9.360793320393483e-07,
+      "loss": 0.1514,
+      "step": 3237
+    },
+    {
+      "epoch": 0.7295462866477033,
+      "grad_norm": 0.4872571475285826,
+      "learning_rate": 9.346263896138e-07,
+      "loss": 0.1531,
+      "step": 3238
+    },
+    {
+      "epoch": 0.7297715943335117,
+      "grad_norm": 0.49925623005616143,
+      "learning_rate": 9.33174316381869e-07,
+      "loss": 0.1521,
+      "step": 3239
+    },
+    {
+      "epoch": 0.7299969020193201,
+      "grad_norm": 0.47482589032112493,
+      "learning_rate": 9.317231131498383e-07,
+      "loss": 0.1381,
+      "step": 3240
+    },
+    {
+      "epoch": 0.7302222097051285,
+      "grad_norm": 0.47642699027713736,
+      "learning_rate": 9.302727807235079e-07,
+      "loss": 0.1583,
+      "step": 3241
+    },
+    {
+      "epoch": 0.730447517390937,
+      "grad_norm": 0.5044203986372681,
+      "learning_rate": 9.288233199081914e-07,
+      "loss": 0.1646,
+      "step": 3242
+    },
+    {
+      "epoch": 0.7306728250767455,
+      "grad_norm": 0.5059155038015059,
+      "learning_rate": 9.273747315087223e-07,
+      "loss": 0.1577,
+      "step": 3243
+    },
+    {
+      "epoch": 0.7308981327625539,
+      "grad_norm": 0.4891834523919038,
+      "learning_rate": 9.259270163294457e-07,
+      "loss": 0.1558,
+      "step": 3244
+    },
+    {
+      "epoch": 0.7311234404483623,
+      "grad_norm": 0.4633981735737557,
+      "learning_rate": 9.244801751742258e-07,
+      "loss": 0.1365,
+      "step": 3245
+    },
+    {
+      "epoch": 0.7313487481341707,
+      "grad_norm": 0.4900172468624769,
+      "learning_rate": 9.230342088464381e-07,
+      "loss": 0.1487,
+      "step": 3246
+    },
+    {
+      "epoch": 0.7315740558199791,
+      "grad_norm": 0.5164669604278073,
+      "learning_rate": 9.215891181489742e-07,
+      "loss": 0.1608,
+      "step": 3247
+    },
+    {
+      "epoch": 0.7317993635057876,
+      "grad_norm": 0.47610476192112533,
+      "learning_rate": 9.201449038842403e-07,
+      "loss": 0.1542,
+      "step": 3248
+    },
+    {
+      "epoch": 0.732024671191596,
+      "grad_norm": 0.4549458362963086,
+      "learning_rate": 9.187015668541526e-07,
+      "loss": 0.1473,
+      "step": 3249
+    },
+    {
+      "epoch": 0.7322499788774045,
+      "grad_norm": 0.48625908990100747,
+      "learning_rate": 9.172591078601448e-07,
+      "loss": 0.1504,
+      "step": 3250
+    },
+    {
+      "epoch": 0.7324752865632129,
+      "grad_norm": 0.4926262365479395,
+      "learning_rate": 9.158175277031584e-07,
+      "loss": 0.1512,
+      "step": 3251
+    },
+    {
+      "epoch": 0.7327005942490213,
+      "grad_norm": 0.5063579543255717,
+      "learning_rate": 9.143768271836506e-07,
+      "loss": 0.1553,
+      "step": 3252
+    },
+    {
+      "epoch": 0.7329259019348298,
+      "grad_norm": 0.4633422630917174,
+      "learning_rate": 9.129370071015886e-07,
+      "loss": 0.1361,
+      "step": 3253
+    },
+    {
+      "epoch": 0.7331512096206382,
+      "grad_norm": 0.4408105126180263,
+      "learning_rate": 9.114980682564492e-07,
+      "loss": 0.1259,
+      "step": 3254
+    },
+    {
+      "epoch": 0.7333765173064466,
+      "grad_norm": 0.5207729527814229,
+      "learning_rate": 9.100600114472238e-07,
+      "loss": 0.1589,
+      "step": 3255
+    },
+    {
+      "epoch": 0.733601824992255,
+      "grad_norm": 0.49598935768061253,
+      "learning_rate": 9.086228374724096e-07,
+      "loss": 0.1538,
+      "step": 3256
+    },
+    {
+      "epoch": 0.7338271326780634,
+      "grad_norm": 0.46713653693216217,
+      "learning_rate": 9.071865471300168e-07,
+      "loss": 0.1513,
+      "step": 3257
+    },
+    {
+      "epoch": 0.734052440363872,
+      "grad_norm": 0.4755809875962297,
+      "learning_rate": 9.057511412175646e-07,
+      "loss": 0.1422,
+      "step": 3258
+    },
+    {
+      "epoch": 0.7342777480496804,
+      "grad_norm": 0.44528902453284236,
+      "learning_rate": 9.043166205320789e-07,
+      "loss": 0.1357,
+      "step": 3259
+    },
+    {
+      "epoch": 0.7345030557354888,
+      "grad_norm": 0.48993203656862394,
+      "learning_rate": 9.028829858700974e-07,
+      "loss": 0.1447,
+      "step": 3260
+    },
+    {
+      "epoch": 0.7347283634212972,
+      "grad_norm": 0.47710624030120524,
+      "learning_rate": 9.014502380276619e-07,
+      "loss": 0.1413,
+      "step": 3261
+    },
+    {
+      "epoch": 0.7349536711071056,
+      "grad_norm": 0.4592113859329267,
+      "learning_rate": 9.000183778003246e-07,
+      "loss": 0.1397,
+      "step": 3262
+    },
+    {
+      "epoch": 0.735178978792914,
+      "grad_norm": 0.48027592981079564,
+      "learning_rate": 8.985874059831456e-07,
+      "loss": 0.1465,
+      "step": 3263
+    },
+    {
+      "epoch": 0.7354042864787225,
+      "grad_norm": 0.4814714122399688,
+      "learning_rate": 8.971573233706881e-07,
+      "loss": 0.1491,
+      "step": 3264
+    },
+    {
+      "epoch": 0.7356295941645309,
+      "grad_norm": 0.4758653935521041,
+      "learning_rate": 8.957281307570254e-07,
+      "loss": 0.1434,
+      "step": 3265
+    },
+    {
+      "epoch": 0.7358549018503394,
+      "grad_norm": 0.4637943976042971,
+      "learning_rate": 8.942998289357333e-07,
+      "loss": 0.1425,
+      "step": 3266
+    },
+    {
+      "epoch": 0.7360802095361478,
+      "grad_norm": 0.44714252120008274,
+      "learning_rate": 8.928724186998961e-07,
+      "loss": 0.1323,
+      "step": 3267
+    },
+    {
+      "epoch": 0.7363055172219563,
+      "grad_norm": 0.5010997935498333,
+      "learning_rate": 8.914459008421e-07,
+      "loss": 0.1555,
+      "step": 3268
+    },
+    {
+      "epoch": 0.7365308249077647,
+      "grad_norm": 0.49517423617843487,
+      "learning_rate": 8.900202761544377e-07,
+      "loss": 0.1466,
+      "step": 3269
+    },
+    {
+      "epoch": 0.7367561325935731,
+      "grad_norm": 0.48746031060642947,
+      "learning_rate": 8.885955454285078e-07,
+      "loss": 0.1463,
+      "step": 3270
+    },
+    {
+      "epoch": 0.7369814402793815,
+      "grad_norm": 0.48248001589906175,
+      "learning_rate": 8.871717094554058e-07,
+      "loss": 0.1444,
+      "step": 3271
+    },
+    {
+      "epoch": 0.7372067479651899,
+      "grad_norm": 0.46149440281114745,
+      "learning_rate": 8.857487690257374e-07,
+      "loss": 0.1314,
+      "step": 3272
+    },
+    {
+      "epoch": 0.7374320556509983,
+      "grad_norm": 0.5005341710829557,
+      "learning_rate": 8.843267249296086e-07,
+      "loss": 0.1482,
+      "step": 3273
+    },
+    {
+      "epoch": 0.7376573633368069,
+      "grad_norm": 0.5015151392851019,
+      "learning_rate": 8.829055779566262e-07,
+      "loss": 0.1512,
+      "step": 3274
+    },
+    {
+      "epoch": 0.7378826710226153,
+      "grad_norm": 0.47516804630917214,
+      "learning_rate": 8.814853288959016e-07,
+      "loss": 0.1488,
+      "step": 3275
+    },
+    {
+      "epoch": 0.7381079787084237,
+      "grad_norm": 0.47292919626104574,
+      "learning_rate": 8.800659785360444e-07,
+      "loss": 0.1503,
+      "step": 3276
+    },
+    {
+      "epoch": 0.7383332863942321,
+      "grad_norm": 0.4899251652160515,
+      "learning_rate": 8.786475276651688e-07,
+      "loss": 0.1421,
+      "step": 3277
+    },
+    {
+      "epoch": 0.7385585940800405,
+      "grad_norm": 0.4691813049653209,
+      "learning_rate": 8.772299770708859e-07,
+      "loss": 0.1251,
+      "step": 3278
+    },
+    {
+      "epoch": 0.738783901765849,
+      "grad_norm": 0.4690066583371361,
+      "learning_rate": 8.758133275403097e-07,
+      "loss": 0.1537,
+      "step": 3279
+    },
+    {
+      "epoch": 0.7390092094516574,
+      "grad_norm": 0.4737472270670779,
+      "learning_rate": 8.743975798600535e-07,
+      "loss": 0.1556,
+      "step": 3280
+    },
+    {
+      "epoch": 0.7392345171374658,
+      "grad_norm": 0.5017685432422245,
+      "learning_rate": 8.729827348162278e-07,
+      "loss": 0.1564,
+      "step": 3281
+    },
+    {
+      "epoch": 0.7394598248232743,
+      "grad_norm": 0.4872577372479285,
+      "learning_rate": 8.71568793194445e-07,
+      "loss": 0.1489,
+      "step": 3282
+    },
+    {
+      "epoch": 0.7396851325090827,
+      "grad_norm": 0.4535495009195591,
+      "learning_rate": 8.701557557798121e-07,
+      "loss": 0.1298,
+      "step": 3283
+    },
+    {
+      "epoch": 0.7399104401948912,
+      "grad_norm": 0.4955801827364163,
+      "learning_rate": 8.687436233569375e-07,
+      "loss": 0.1605,
+      "step": 3284
+    },
+    {
+      "epoch": 0.7401357478806996,
+      "grad_norm": 0.46483112374305074,
+      "learning_rate": 8.673323967099259e-07,
+      "loss": 0.1441,
+      "step": 3285
+    },
+    {
+      "epoch": 0.740361055566508,
+      "grad_norm": 0.48141735604131547,
+      "learning_rate": 8.659220766223778e-07,
+      "loss": 0.1462,
+      "step": 3286
+    },
+    {
+      "epoch": 0.7405863632523164,
+      "grad_norm": 0.4862752455378802,
+      "learning_rate": 8.645126638773926e-07,
+      "loss": 0.153,
+      "step": 3287
+    },
+    {
+      "epoch": 0.7408116709381248,
+      "grad_norm": 0.46082477554977386,
+      "learning_rate": 8.631041592575643e-07,
+      "loss": 0.1339,
+      "step": 3288
+    },
+    {
+      "epoch": 0.7410369786239334,
+      "grad_norm": 0.47467778641399166,
+      "learning_rate": 8.616965635449814e-07,
+      "loss": 0.1545,
+      "step": 3289
+    },
+    {
+      "epoch": 0.7412622863097418,
+      "grad_norm": 0.45990095150736854,
+      "learning_rate": 8.602898775212317e-07,
+      "loss": 0.1356,
+      "step": 3290
+    },
+    {
+      "epoch": 0.7414875939955502,
+      "grad_norm": 0.48722496822548544,
+      "learning_rate": 8.588841019673938e-07,
+      "loss": 0.1455,
+      "step": 3291
+    },
+    {
+      "epoch": 0.7417129016813586,
+      "grad_norm": 0.47337013865038563,
+      "learning_rate": 8.57479237664044e-07,
+      "loss": 0.142,
+      "step": 3292
+    },
+    {
+      "epoch": 0.741938209367167,
+      "grad_norm": 0.5024652414346156,
+      "learning_rate": 8.560752853912494e-07,
+      "loss": 0.1448,
+      "step": 3293
+    },
+    {
+      "epoch": 0.7421635170529755,
+      "grad_norm": 0.4809913420160176,
+      "learning_rate": 8.546722459285727e-07,
+      "loss": 0.15,
+      "step": 3294
+    },
+    {
+      "epoch": 0.7423888247387839,
+      "grad_norm": 0.4999499323219179,
+      "learning_rate": 8.532701200550714e-07,
+      "loss": 0.1481,
+      "step": 3295
+    },
+    {
+      "epoch": 0.7426141324245923,
+      "grad_norm": 0.4689592016416926,
+      "learning_rate": 8.518689085492909e-07,
+      "loss": 0.1441,
+      "step": 3296
+    },
+    {
+      "epoch": 0.7428394401104008,
+      "grad_norm": 0.46670042652287225,
+      "learning_rate": 8.504686121892741e-07,
+      "loss": 0.15,
+      "step": 3297
+    },
+    {
+      "epoch": 0.7430647477962092,
+      "grad_norm": 0.4846860503801346,
+      "learning_rate": 8.490692317525514e-07,
+      "loss": 0.1426,
+      "step": 3298
+    },
+    {
+      "epoch": 0.7432900554820177,
+      "grad_norm": 0.4821622091862189,
+      "learning_rate": 8.476707680161486e-07,
+      "loss": 0.1478,
+      "step": 3299
+    },
+    {
+      "epoch": 0.7435153631678261,
+      "grad_norm": 0.47393658159990576,
+      "learning_rate": 8.462732217565783e-07,
+      "loss": 0.142,
+      "step": 3300
+    },
+    {
+      "epoch": 0.7437406708536345,
+      "grad_norm": 0.4800748838632915,
+      "learning_rate": 8.448765937498471e-07,
+      "loss": 0.147,
+      "step": 3301
+    },
+    {
+      "epoch": 0.7439659785394429,
+      "grad_norm": 0.507382353930732,
+      "learning_rate": 8.434808847714512e-07,
+      "loss": 0.1501,
+      "step": 3302
+    },
+    {
+      "epoch": 0.7441912862252513,
+      "grad_norm": 0.47714205294667555,
+      "learning_rate": 8.420860955963739e-07,
+      "loss": 0.1503,
+      "step": 3303
+    },
+    {
+      "epoch": 0.7444165939110597,
+      "grad_norm": 0.5072068700018229,
+      "learning_rate": 8.406922269990917e-07,
+      "loss": 0.1615,
+      "step": 3304
+    },
+    {
+      "epoch": 0.7446419015968683,
+      "grad_norm": 0.4982128659213873,
+      "learning_rate": 8.392992797535666e-07,
+      "loss": 0.1512,
+      "step": 3305
+    },
+    {
+      "epoch": 0.7448672092826767,
+      "grad_norm": 0.46834497278773835,
+      "learning_rate": 8.379072546332498e-07,
+      "loss": 0.1386,
+      "step": 3306
+    },
+    {
+      "epoch": 0.7450925169684851,
+      "grad_norm": 0.472453143243406,
+      "learning_rate": 8.365161524110823e-07,
+      "loss": 0.1431,
+      "step": 3307
+    },
+    {
+      "epoch": 0.7453178246542935,
+      "grad_norm": 0.46279948878360144,
+      "learning_rate": 8.351259738594902e-07,
+      "loss": 0.1479,
+      "step": 3308
+    },
+    {
+      "epoch": 0.7455431323401019,
+      "grad_norm": 0.4627044576661592,
+      "learning_rate": 8.337367197503881e-07,
+      "loss": 0.14,
+      "step": 3309
+    },
+    {
+      "epoch": 0.7457684400259104,
+      "grad_norm": 0.4819754464823401,
+      "learning_rate": 8.323483908551783e-07,
+      "loss": 0.1452,
+      "step": 3310
+    },
+    {
+      "epoch": 0.7459937477117188,
+      "grad_norm": 0.4716373648789507,
+      "learning_rate": 8.309609879447458e-07,
+      "loss": 0.1439,
+      "step": 3311
+    },
+    {
+      "epoch": 0.7462190553975272,
+      "grad_norm": 0.43775940967744453,
+      "learning_rate": 8.29574511789466e-07,
+      "loss": 0.1339,
+      "step": 3312
+    },
+    {
+      "epoch": 0.7464443630833357,
+      "grad_norm": 0.4797136707395238,
+      "learning_rate": 8.281889631591955e-07,
+      "loss": 0.154,
+      "step": 3313
+    },
+    {
+      "epoch": 0.7466696707691441,
+      "grad_norm": 0.4896839562641403,
+      "learning_rate": 8.268043428232798e-07,
+      "loss": 0.162,
+      "step": 3314
+    },
+    {
+      "epoch": 0.7468949784549526,
+      "grad_norm": 0.46071254712821796,
+      "learning_rate": 8.254206515505444e-07,
+      "loss": 0.1453,
+      "step": 3315
+    },
+    {
+      "epoch": 0.747120286140761,
+      "grad_norm": 0.45393470832573807,
+      "learning_rate": 8.240378901093035e-07,
+      "loss": 0.1347,
+      "step": 3316
+    },
+    {
+      "epoch": 0.7473455938265694,
+      "grad_norm": 0.4647022595151276,
+      "learning_rate": 8.22656059267353e-07,
+      "loss": 0.1367,
+      "step": 3317
+    },
+    {
+      "epoch": 0.7475709015123778,
+      "grad_norm": 0.4894365793407213,
+      "learning_rate": 8.212751597919708e-07,
+      "loss": 0.1463,
+      "step": 3318
+    },
+    {
+      "epoch": 0.7477962091981862,
+      "grad_norm": 0.4874094265856785,
+      "learning_rate": 8.198951924499202e-07,
+      "loss": 0.1533,
+      "step": 3319
+    },
+    {
+      "epoch": 0.7480215168839947,
+      "grad_norm": 0.4552551653740835,
+      "learning_rate": 8.185161580074444e-07,
+      "loss": 0.1367,
+      "step": 3320
+    },
+    {
+      "epoch": 0.7482468245698032,
+      "grad_norm": 0.4676081851330181,
+      "learning_rate": 8.171380572302712e-07,
+      "loss": 0.1378,
+      "step": 3321
+    },
+    {
+      "epoch": 0.7484721322556116,
+      "grad_norm": 0.4982658730940427,
+      "learning_rate": 8.157608908836071e-07,
+      "loss": 0.1536,
+      "step": 3322
+    },
+    {
+      "epoch": 0.74869743994142,
+      "grad_norm": 0.4569590223747696,
+      "learning_rate": 8.143846597321425e-07,
+      "loss": 0.1341,
+      "step": 3323
+    },
+    {
+      "epoch": 0.7489227476272284,
+      "grad_norm": 0.507077702702224,
+      "learning_rate": 8.130093645400469e-07,
+      "loss": 0.1631,
+      "step": 3324
+    },
+    {
+      "epoch": 0.7491480553130369,
+      "grad_norm": 0.4800642788710456,
+      "learning_rate": 8.116350060709696e-07,
+      "loss": 0.1461,
+      "step": 3325
+    },
+    {
+      "epoch": 0.7493733629988453,
+      "grad_norm": 0.4446086202012074,
+      "learning_rate": 8.102615850880413e-07,
+      "loss": 0.1246,
+      "step": 3326
+    },
+    {
+      "epoch": 0.7495986706846537,
+      "grad_norm": 0.4562022923162509,
+      "learning_rate": 8.088891023538722e-07,
+      "loss": 0.1303,
+      "step": 3327
+    },
+    {
+      "epoch": 0.7498239783704621,
+      "grad_norm": 0.5165662372340282,
+      "learning_rate": 8.075175586305492e-07,
+      "loss": 0.1656,
+      "step": 3328
+    },
+    {
+      "epoch": 0.7500492860562706,
+      "grad_norm": 0.5027656419852117,
+      "learning_rate": 8.061469546796413e-07,
+      "loss": 0.1646,
+      "step": 3329
+    },
+    {
+      "epoch": 0.750274593742079,
+      "grad_norm": 0.47424795152236715,
+      "learning_rate": 8.047772912621921e-07,
+      "loss": 0.147,
+      "step": 3330
+    },
+    {
+      "epoch": 0.7504999014278875,
+      "grad_norm": 0.4799229558407996,
+      "learning_rate": 8.034085691387253e-07,
+      "loss": 0.1554,
+      "step": 3331
+    },
+    {
+      "epoch": 0.7507252091136959,
+      "grad_norm": 0.44539800876812674,
+      "learning_rate": 8.020407890692419e-07,
+      "loss": 0.1284,
+      "step": 3332
+    },
+    {
+      "epoch": 0.7509505167995043,
+      "grad_norm": 0.46596737623628415,
+      "learning_rate": 8.006739518132179e-07,
+      "loss": 0.1373,
+      "step": 3333
+    },
+    {
+      "epoch": 0.7511758244853127,
+      "grad_norm": 0.47852726088194664,
+      "learning_rate": 7.993080581296087e-07,
+      "loss": 0.1469,
+      "step": 3334
+    },
+    {
+      "epoch": 0.7514011321711211,
+      "grad_norm": 0.46838224015083574,
+      "learning_rate": 7.979431087768424e-07,
+      "loss": 0.1377,
+      "step": 3335
+    },
+    {
+      "epoch": 0.7516264398569297,
+      "grad_norm": 0.487296225417356,
+      "learning_rate": 7.96579104512826e-07,
+      "loss": 0.1427,
+      "step": 3336
+    },
+    {
+      "epoch": 0.7518517475427381,
+      "grad_norm": 0.46682500251151077,
+      "learning_rate": 7.95216046094939e-07,
+      "loss": 0.1379,
+      "step": 3337
+    },
+    {
+      "epoch": 0.7520770552285465,
+      "grad_norm": 0.48515341220916963,
+      "learning_rate": 7.938539342800373e-07,
+      "loss": 0.1438,
+      "step": 3338
+    },
+    {
+      "epoch": 0.7523023629143549,
+      "grad_norm": 0.46473632519103775,
+      "learning_rate": 7.92492769824452e-07,
+      "loss": 0.1334,
+      "step": 3339
+    },
+    {
+      "epoch": 0.7525276706001633,
+      "grad_norm": 0.4719673542349233,
+      "learning_rate": 7.911325534839851e-07,
+      "loss": 0.1368,
+      "step": 3340
+    },
+    {
+      "epoch": 0.7527529782859718,
+      "grad_norm": 0.47032858945280154,
+      "learning_rate": 7.897732860139157e-07,
+      "loss": 0.1367,
+      "step": 3341
+    },
+    {
+      "epoch": 0.7529782859717802,
+      "grad_norm": 0.45765166719265943,
+      "learning_rate": 7.884149681689937e-07,
+      "loss": 0.1376,
+      "step": 3342
+    },
+    {
+      "epoch": 0.7532035936575886,
+      "grad_norm": 0.47212795628146054,
+      "learning_rate": 7.870576007034414e-07,
+      "loss": 0.1424,
+      "step": 3343
+    },
+    {
+      "epoch": 0.7534289013433971,
+      "grad_norm": 0.47987368156781673,
+      "learning_rate": 7.857011843709559e-07,
+      "loss": 0.1429,
+      "step": 3344
+    },
+    {
+      "epoch": 0.7536542090292055,
+      "grad_norm": 0.5245624120355394,
+      "learning_rate": 7.843457199247034e-07,
+      "loss": 0.1566,
+      "step": 3345
+    },
+    {
+      "epoch": 0.753879516715014,
+      "grad_norm": 0.4965488224022442,
+      "learning_rate": 7.829912081173238e-07,
+      "loss": 0.1427,
+      "step": 3346
+    },
+    {
+      "epoch": 0.7541048244008224,
+      "grad_norm": 0.47706440489835394,
+      "learning_rate": 7.816376497009262e-07,
+      "loss": 0.1335,
+      "step": 3347
+    },
+    {
+      "epoch": 0.7543301320866308,
+      "grad_norm": 0.48299077652397493,
+      "learning_rate": 7.802850454270913e-07,
+      "loss": 0.1449,
+      "step": 3348
+    },
+    {
+      "epoch": 0.7545554397724392,
+      "grad_norm": 0.4741612790797712,
+      "learning_rate": 7.789333960468707e-07,
+      "loss": 0.1445,
+      "step": 3349
+    },
+    {
+      "epoch": 0.7547807474582476,
+      "grad_norm": 0.49498587339603634,
+      "learning_rate": 7.775827023107835e-07,
+      "loss": 0.154,
+      "step": 3350
+    },
+    {
+      "epoch": 0.755006055144056,
+      "grad_norm": 0.46834623356511995,
+      "learning_rate": 7.762329649688214e-07,
+      "loss": 0.1353,
+      "step": 3351
+    },
+    {
+      "epoch": 0.7552313628298646,
+      "grad_norm": 0.4661163814611361,
+      "learning_rate": 7.74884184770441e-07,
+      "loss": 0.1292,
+      "step": 3352
+    },
+    {
+      "epoch": 0.755456670515673,
+      "grad_norm": 0.4981050040486079,
+      "learning_rate": 7.735363624645712e-07,
+      "loss": 0.1477,
+      "step": 3353
+    },
+    {
+      "epoch": 0.7556819782014814,
+      "grad_norm": 0.4480961010367873,
+      "learning_rate": 7.721894987996076e-07,
+      "loss": 0.1363,
+      "step": 3354
+    },
+    {
+      "epoch": 0.7559072858872898,
+      "grad_norm": 0.47967498498134536,
+      "learning_rate": 7.708435945234124e-07,
+      "loss": 0.1431,
+      "step": 3355
+    },
+    {
+      "epoch": 0.7561325935730983,
+      "grad_norm": 0.4670216916507628,
+      "learning_rate": 7.694986503833171e-07,
+      "loss": 0.1479,
+      "step": 3356
+    },
+    {
+      "epoch": 0.7563579012589067,
+      "grad_norm": 0.515124251928348,
+      "learning_rate": 7.681546671261181e-07,
+      "loss": 0.1624,
+      "step": 3357
+    },
+    {
+      "epoch": 0.7565832089447151,
+      "grad_norm": 0.5047300539541966,
+      "learning_rate": 7.668116454980804e-07,
+      "loss": 0.1455,
+      "step": 3358
+    },
+    {
+      "epoch": 0.7568085166305235,
+      "grad_norm": 0.5022582248887345,
+      "learning_rate": 7.654695862449327e-07,
+      "loss": 0.1563,
+      "step": 3359
+    },
+    {
+      "epoch": 0.757033824316332,
+      "grad_norm": 0.4403365656908437,
+      "learning_rate": 7.641284901118703e-07,
+      "loss": 0.1241,
+      "step": 3360
+    },
+    {
+      "epoch": 0.7572591320021405,
+      "grad_norm": 0.46478175233204677,
+      "learning_rate": 7.627883578435555e-07,
+      "loss": 0.1495,
+      "step": 3361
+    },
+    {
+      "epoch": 0.7574844396879489,
+      "grad_norm": 0.4680079929371967,
+      "learning_rate": 7.614491901841118e-07,
+      "loss": 0.1385,
+      "step": 3362
+    },
+    {
+      "epoch": 0.7577097473737573,
+      "grad_norm": 0.45630374777367133,
+      "learning_rate": 7.601109878771301e-07,
+      "loss": 0.1404,
+      "step": 3363
+    },
+    {
+      "epoch": 0.7579350550595657,
+      "grad_norm": 0.48047432753197256,
+      "learning_rate": 7.587737516656651e-07,
+      "loss": 0.1523,
+      "step": 3364
+    },
+    {
+      "epoch": 0.7581603627453741,
+      "grad_norm": 0.4754340021655358,
+      "learning_rate": 7.574374822922323e-07,
+      "loss": 0.146,
+      "step": 3365
+    },
+    {
+      "epoch": 0.7583856704311825,
+      "grad_norm": 0.4623116504838362,
+      "learning_rate": 7.561021804988141e-07,
+      "loss": 0.1314,
+      "step": 3366
+    },
+    {
+      "epoch": 0.758610978116991,
+      "grad_norm": 0.4804463957409756,
+      "learning_rate": 7.547678470268526e-07,
+      "loss": 0.1458,
+      "step": 3367
+    },
+    {
+      "epoch": 0.7588362858027995,
+      "grad_norm": 0.4629048572883021,
+      "learning_rate": 7.534344826172546e-07,
+      "loss": 0.1378,
+      "step": 3368
+    },
+    {
+      "epoch": 0.7590615934886079,
+      "grad_norm": 0.48475687389213085,
+      "learning_rate": 7.52102088010386e-07,
+      "loss": 0.1471,
+      "step": 3369
+    },
+    {
+      "epoch": 0.7592869011744163,
+      "grad_norm": 0.43776274622028544,
+      "learning_rate": 7.507706639460768e-07,
+      "loss": 0.1194,
+      "step": 3370
+    },
+    {
+      "epoch": 0.7595122088602247,
+      "grad_norm": 0.49795081371303823,
+      "learning_rate": 7.494402111636179e-07,
+      "loss": 0.1586,
+      "step": 3371
+    },
+    {
+      "epoch": 0.7597375165460332,
+      "grad_norm": 0.4772160878473002,
+      "learning_rate": 7.481107304017588e-07,
+      "loss": 0.1391,
+      "step": 3372
+    },
+    {
+      "epoch": 0.7599628242318416,
+      "grad_norm": 0.4881633011030423,
+      "learning_rate": 7.467822223987117e-07,
+      "loss": 0.1468,
+      "step": 3373
+    },
+    {
+      "epoch": 0.76018813191765,
+      "grad_norm": 0.4832816618224351,
+      "learning_rate": 7.454546878921465e-07,
+      "loss": 0.1383,
+      "step": 3374
+    },
+    {
+      "epoch": 0.7604134396034584,
+      "grad_norm": 0.4553465039777067,
+      "learning_rate": 7.441281276191939e-07,
+      "loss": 0.1299,
+      "step": 3375
+    },
+    {
+      "epoch": 0.760638747289267,
+      "grad_norm": 0.4563435010027247,
+      "learning_rate": 7.428025423164456e-07,
+      "loss": 0.1317,
+      "step": 3376
+    },
+    {
+      "epoch": 0.7608640549750754,
+      "grad_norm": 0.49239236621925697,
+      "learning_rate": 7.414779327199461e-07,
+      "loss": 0.1547,
+      "step": 3377
+    },
+    {
+      "epoch": 0.7610893626608838,
+      "grad_norm": 0.501508390358165,
+      "learning_rate": 7.401542995652033e-07,
+      "loss": 0.1506,
+      "step": 3378
+    },
+    {
+      "epoch": 0.7613146703466922,
+      "grad_norm": 0.48179929137762706,
+      "learning_rate": 7.388316435871825e-07,
+      "loss": 0.1404,
+      "step": 3379
+    },
+    {
+      "epoch": 0.7615399780325006,
+      "grad_norm": 0.4628844742100713,
+      "learning_rate": 7.375099655203033e-07,
+      "loss": 0.1369,
+      "step": 3380
+    },
+    {
+      "epoch": 0.761765285718309,
+      "grad_norm": 0.4575415724895742,
+      "learning_rate": 7.361892660984459e-07,
+      "loss": 0.134,
+      "step": 3381
+    },
+    {
+      "epoch": 0.7619905934041175,
+      "grad_norm": 0.4445676062794546,
+      "learning_rate": 7.348695460549443e-07,
+      "loss": 0.132,
+      "step": 3382
+    },
+    {
+      "epoch": 0.762215901089926,
+      "grad_norm": 0.48156865787255837,
+      "learning_rate": 7.335508061225907e-07,
+      "loss": 0.1482,
+      "step": 3383
+    },
+    {
+      "epoch": 0.7624412087757344,
+      "grad_norm": 0.4722120592143978,
+      "learning_rate": 7.322330470336314e-07,
+      "loss": 0.1487,
+      "step": 3384
+    },
+    {
+      "epoch": 0.7626665164615428,
+      "grad_norm": 0.45923459135561195,
+      "learning_rate": 7.309162695197692e-07,
+      "loss": 0.1401,
+      "step": 3385
+    },
+    {
+      "epoch": 0.7628918241473512,
+      "grad_norm": 0.49513812336272167,
+      "learning_rate": 7.296004743121627e-07,
+      "loss": 0.1434,
+      "step": 3386
+    },
+    {
+      "epoch": 0.7631171318331597,
+      "grad_norm": 0.47090809838121395,
+      "learning_rate": 7.28285662141422e-07,
+      "loss": 0.137,
+      "step": 3387
+    },
+    {
+      "epoch": 0.7633424395189681,
+      "grad_norm": 0.48067072748376605,
+      "learning_rate": 7.26971833737615e-07,
+      "loss": 0.1552,
+      "step": 3388
+    },
+    {
+      "epoch": 0.7635677472047765,
+      "grad_norm": 0.49326359809106446,
+      "learning_rate": 7.256589898302599e-07,
+      "loss": 0.1421,
+      "step": 3389
+    },
+    {
+      "epoch": 0.7637930548905849,
+      "grad_norm": 0.4709260975136618,
+      "learning_rate": 7.243471311483322e-07,
+      "loss": 0.1373,
+      "step": 3390
+    },
+    {
+      "epoch": 0.7640183625763934,
+      "grad_norm": 0.46602218525293615,
+      "learning_rate": 7.230362584202557e-07,
+      "loss": 0.1375,
+      "step": 3391
+    },
+    {
+      "epoch": 0.7642436702622019,
+      "grad_norm": 0.488504880542589,
+      "learning_rate": 7.217263723739107e-07,
+      "loss": 0.1434,
+      "step": 3392
+    },
+    {
+      "epoch": 0.7644689779480103,
+      "grad_norm": 0.45585994244897526,
+      "learning_rate": 7.204174737366293e-07,
+      "loss": 0.1358,
+      "step": 3393
+    },
+    {
+      "epoch": 0.7646942856338187,
+      "grad_norm": 0.4584522893893621,
+      "learning_rate": 7.191095632351908e-07,
+      "loss": 0.1428,
+      "step": 3394
+    },
+    {
+      "epoch": 0.7649195933196271,
+      "grad_norm": 0.4968819934814164,
+      "learning_rate": 7.178026415958311e-07,
+      "loss": 0.1487,
+      "step": 3395
+    },
+    {
+      "epoch": 0.7651449010054355,
+      "grad_norm": 0.5065663037710753,
+      "learning_rate": 7.164967095442357e-07,
+      "loss": 0.163,
+      "step": 3396
+    },
+    {
+      "epoch": 0.765370208691244,
+      "grad_norm": 0.487232527516781,
+      "learning_rate": 7.151917678055384e-07,
+      "loss": 0.1492,
+      "step": 3397
+    },
+    {
+      "epoch": 0.7655955163770524,
+      "grad_norm": 0.4678617657264611,
+      "learning_rate": 7.138878171043262e-07,
+      "loss": 0.1381,
+      "step": 3398
+    },
+    {
+      "epoch": 0.7658208240628609,
+      "grad_norm": 0.49173600814669405,
+      "learning_rate": 7.125848581646327e-07,
+      "loss": 0.1566,
+      "step": 3399
+    },
+    {
+      "epoch": 0.7660461317486693,
+      "grad_norm": 0.4726487112124558,
+      "learning_rate": 7.112828917099438e-07,
+      "loss": 0.137,
+      "step": 3400
+    },
+    {
+      "epoch": 0.7662714394344777,
+      "grad_norm": 0.4680839672704657,
+      "learning_rate": 7.099819184631929e-07,
+      "loss": 0.1407,
+      "step": 3401
+    },
+    {
+      "epoch": 0.7664967471202861,
+      "grad_norm": 0.4714623274461953,
+      "learning_rate": 7.086819391467612e-07,
+      "loss": 0.1445,
+      "step": 3402
+    },
+    {
+      "epoch": 0.7667220548060946,
+      "grad_norm": 0.4985792687519659,
+      "learning_rate": 7.073829544824795e-07,
+      "loss": 0.1473,
+      "step": 3403
+    },
+    {
+      "epoch": 0.766947362491903,
+      "grad_norm": 0.46973950135993436,
+      "learning_rate": 7.060849651916244e-07,
+      "loss": 0.1389,
+      "step": 3404
+    },
+    {
+      "epoch": 0.7671726701777114,
+      "grad_norm": 0.48334204954281296,
+      "learning_rate": 7.047879719949227e-07,
+      "loss": 0.1414,
+      "step": 3405
+    },
+    {
+      "epoch": 0.7673979778635198,
+      "grad_norm": 0.46175218020844494,
+      "learning_rate": 7.034919756125447e-07,
+      "loss": 0.1394,
+      "step": 3406
+    },
+    {
+      "epoch": 0.7676232855493283,
+      "grad_norm": 0.49410571707078754,
+      "learning_rate": 7.021969767641096e-07,
+      "loss": 0.1423,
+      "step": 3407
+    },
+    {
+      "epoch": 0.7678485932351368,
+      "grad_norm": 0.501188825151817,
+      "learning_rate": 7.009029761686825e-07,
+      "loss": 0.1496,
+      "step": 3408
+    },
+    {
+      "epoch": 0.7680739009209452,
+      "grad_norm": 0.4793266257229126,
+      "learning_rate": 6.996099745447726e-07,
+      "loss": 0.1407,
+      "step": 3409
+    },
+    {
+      "epoch": 0.7682992086067536,
+      "grad_norm": 0.5061394233782135,
+      "learning_rate": 6.98317972610337e-07,
+      "loss": 0.1578,
+      "step": 3410
+    },
+    {
+      "epoch": 0.768524516292562,
+      "grad_norm": 0.4338884896826393,
+      "learning_rate": 6.970269710827754e-07,
+      "loss": 0.1235,
+      "step": 3411
+    },
+    {
+      "epoch": 0.7687498239783704,
+      "grad_norm": 0.45937831755042746,
+      "learning_rate": 6.957369706789319e-07,
+      "loss": 0.143,
+      "step": 3412
+    },
+    {
+      "epoch": 0.7689751316641789,
+      "grad_norm": 0.47034142663055073,
+      "learning_rate": 6.944479721150971e-07,
+      "loss": 0.1359,
+      "step": 3413
+    },
+    {
+      "epoch": 0.7692004393499873,
+      "grad_norm": 0.4960066484738008,
+      "learning_rate": 6.931599761070027e-07,
+      "loss": 0.1539,
+      "step": 3414
+    },
+    {
+      "epoch": 0.7694257470357958,
+      "grad_norm": 0.4709523524894798,
+      "learning_rate": 6.91872983369826e-07,
+      "loss": 0.1504,
+      "step": 3415
+    },
+    {
+      "epoch": 0.7696510547216042,
+      "grad_norm": 0.47377291373340263,
+      "learning_rate": 6.905869946181848e-07,
+      "loss": 0.1488,
+      "step": 3416
+    },
+    {
+      "epoch": 0.7698763624074126,
+      "grad_norm": 0.47037382567724695,
+      "learning_rate": 6.893020105661416e-07,
+      "loss": 0.1397,
+      "step": 3417
+    },
+    {
+      "epoch": 0.7701016700932211,
+      "grad_norm": 0.4685633998550794,
+      "learning_rate": 6.880180319272006e-07,
+      "loss": 0.1419,
+      "step": 3418
+    },
+    {
+      "epoch": 0.7703269777790295,
+      "grad_norm": 0.4907543802869892,
+      "learning_rate": 6.867350594143058e-07,
+      "loss": 0.1568,
+      "step": 3419
+    },
+    {
+      "epoch": 0.7705522854648379,
+      "grad_norm": 0.49679497792993677,
+      "learning_rate": 6.854530937398459e-07,
+      "loss": 0.1487,
+      "step": 3420
+    },
+    {
+      "epoch": 0.7707775931506463,
+      "grad_norm": 0.4654299885121034,
+      "learning_rate": 6.841721356156466e-07,
+      "loss": 0.1495,
+      "step": 3421
+    },
+    {
+      "epoch": 0.7710029008364547,
+      "grad_norm": 0.4374517267903433,
+      "learning_rate": 6.828921857529774e-07,
+      "loss": 0.1311,
+      "step": 3422
+    },
+    {
+      "epoch": 0.7712282085222633,
+      "grad_norm": 0.4597646748074688,
+      "learning_rate": 6.816132448625474e-07,
+      "loss": 0.1373,
+      "step": 3423
+    },
+    {
+      "epoch": 0.7714535162080717,
+      "grad_norm": 0.47486822261268397,
+      "learning_rate": 6.803353136545033e-07,
+      "loss": 0.1399,
+      "step": 3424
+    },
+    {
+      "epoch": 0.7716788238938801,
+      "grad_norm": 0.48137920510488824,
+      "learning_rate": 6.790583928384339e-07,
+      "loss": 0.1492,
+      "step": 3425
+    },
+    {
+      "epoch": 0.7719041315796885,
+      "grad_norm": 0.5092249833274699,
+      "learning_rate": 6.777824831233645e-07,
+      "loss": 0.156,
+      "step": 3426
+    },
+    {
+      "epoch": 0.7721294392654969,
+      "grad_norm": 0.4791475636794947,
+      "learning_rate": 6.765075852177619e-07,
+      "loss": 0.1417,
+      "step": 3427
+    },
+    {
+      "epoch": 0.7723547469513053,
+      "grad_norm": 0.49452899112181636,
+      "learning_rate": 6.752336998295281e-07,
+      "loss": 0.1488,
+      "step": 3428
+    },
+    {
+      "epoch": 0.7725800546371138,
+      "grad_norm": 0.48047913950608484,
+      "learning_rate": 6.739608276660037e-07,
+      "loss": 0.147,
+      "step": 3429
+    },
+    {
+      "epoch": 0.7728053623229223,
+      "grad_norm": 0.4724754669670758,
+      "learning_rate": 6.726889694339689e-07,
+      "loss": 0.1433,
+      "step": 3430
+    },
+    {
+      "epoch": 0.7730306700087307,
+      "grad_norm": 0.503636873558274,
+      "learning_rate": 6.714181258396371e-07,
+      "loss": 0.1652,
+      "step": 3431
+    },
+    {
+      "epoch": 0.7732559776945391,
+      "grad_norm": 0.50474994731612,
+      "learning_rate": 6.701482975886617e-07,
+      "loss": 0.1597,
+      "step": 3432
+    },
+    {
+      "epoch": 0.7734812853803475,
+      "grad_norm": 0.5076282017497796,
+      "learning_rate": 6.688794853861316e-07,
+      "loss": 0.1538,
+      "step": 3433
+    },
+    {
+      "epoch": 0.773706593066156,
+      "grad_norm": 0.48011437833845344,
+      "learning_rate": 6.676116899365692e-07,
+      "loss": 0.1463,
+      "step": 3434
+    },
+    {
+      "epoch": 0.7739319007519644,
+      "grad_norm": 0.4666723449374404,
+      "learning_rate": 6.663449119439358e-07,
+      "loss": 0.1304,
+      "step": 3435
+    },
+    {
+      "epoch": 0.7741572084377728,
+      "grad_norm": 0.47606931232813704,
+      "learning_rate": 6.650791521116243e-07,
+      "loss": 0.1344,
+      "step": 3436
+    },
+    {
+      "epoch": 0.7743825161235812,
+      "grad_norm": 0.4792947664321174,
+      "learning_rate": 6.638144111424655e-07,
+      "loss": 0.1448,
+      "step": 3437
+    },
+    {
+      "epoch": 0.7746078238093898,
+      "grad_norm": 0.483992588918863,
+      "learning_rate": 6.625506897387215e-07,
+      "loss": 0.1347,
+      "step": 3438
+    },
+    {
+      "epoch": 0.7748331314951982,
+      "grad_norm": 0.49226446774811605,
+      "learning_rate": 6.612879886020907e-07,
+      "loss": 0.1524,
+      "step": 3439
+    },
+    {
+      "epoch": 0.7750584391810066,
+      "grad_norm": 0.4876555810986814,
+      "learning_rate": 6.600263084337041e-07,
+      "loss": 0.1416,
+      "step": 3440
+    },
+    {
+      "epoch": 0.775283746866815,
+      "grad_norm": 0.4879932793568683,
+      "learning_rate": 6.587656499341247e-07,
+      "loss": 0.1485,
+      "step": 3441
+    },
+    {
+      "epoch": 0.7755090545526234,
+      "grad_norm": 0.4637289855714098,
+      "learning_rate": 6.575060138033504e-07,
+      "loss": 0.1369,
+      "step": 3442
+    },
+    {
+      "epoch": 0.7757343622384318,
+      "grad_norm": 0.462783115914039,
+      "learning_rate": 6.562474007408087e-07,
+      "loss": 0.1403,
+      "step": 3443
+    },
+    {
+      "epoch": 0.7759596699242403,
+      "grad_norm": 0.45701063708272344,
+      "learning_rate": 6.549898114453615e-07,
+      "loss": 0.1356,
+      "step": 3444
+    },
+    {
+      "epoch": 0.7761849776100487,
+      "grad_norm": 0.4885466428022282,
+      "learning_rate": 6.537332466153018e-07,
+      "loss": 0.1458,
+      "step": 3445
+    },
+    {
+      "epoch": 0.7764102852958572,
+      "grad_norm": 0.4629424193330636,
+      "learning_rate": 6.524777069483526e-07,
+      "loss": 0.1437,
+      "step": 3446
+    },
+    {
+      "epoch": 0.7766355929816656,
+      "grad_norm": 0.5012797628936799,
+      "learning_rate": 6.512231931416674e-07,
+      "loss": 0.1646,
+      "step": 3447
+    },
+    {
+      "epoch": 0.776860900667474,
+      "grad_norm": 0.4904692323067029,
+      "learning_rate": 6.499697058918326e-07,
+      "loss": 0.1428,
+      "step": 3448
+    },
+    {
+      "epoch": 0.7770862083532825,
+      "grad_norm": 0.4813223757776377,
+      "learning_rate": 6.487172458948612e-07,
+      "loss": 0.1489,
+      "step": 3449
+    },
+    {
+      "epoch": 0.7773115160390909,
+      "grad_norm": 0.49212730316398134,
+      "learning_rate": 6.474658138461992e-07,
+      "loss": 0.1523,
+      "step": 3450
+    },
+    {
+      "epoch": 0.7775368237248993,
+      "grad_norm": 0.4834210288352534,
+      "learning_rate": 6.462154104407187e-07,
+      "loss": 0.1464,
+      "step": 3451
+    },
+    {
+      "epoch": 0.7777621314107077,
+      "grad_norm": 0.44844795640795554,
+      "learning_rate": 6.449660363727236e-07,
+      "loss": 0.1278,
+      "step": 3452
+    },
+    {
+      "epoch": 0.7779874390965161,
+      "grad_norm": 0.4713579351940081,
+      "learning_rate": 6.437176923359434e-07,
+      "loss": 0.1378,
+      "step": 3453
+    },
+    {
+      "epoch": 0.7782127467823247,
+      "grad_norm": 0.5120717655518697,
+      "learning_rate": 6.424703790235374e-07,
+      "loss": 0.1464,
+      "step": 3454
+    },
+    {
+      "epoch": 0.7784380544681331,
+      "grad_norm": 0.48508407218751737,
+      "learning_rate": 6.41224097128093e-07,
+      "loss": 0.146,
+      "step": 3455
+    },
+    {
+      "epoch": 0.7786633621539415,
+      "grad_norm": 0.5042237917237723,
+      "learning_rate": 6.399788473416229e-07,
+      "loss": 0.1628,
+      "step": 3456
+    },
+    {
+      "epoch": 0.7788886698397499,
+      "grad_norm": 0.491435891395647,
+      "learning_rate": 6.387346303555691e-07,
+      "loss": 0.1456,
+      "step": 3457
+    },
+    {
+      "epoch": 0.7791139775255583,
+      "grad_norm": 0.5030006112904724,
+      "learning_rate": 6.374914468607976e-07,
+      "loss": 0.1483,
+      "step": 3458
+    },
+    {
+      "epoch": 0.7793392852113668,
+      "grad_norm": 0.4831057771694969,
+      "learning_rate": 6.362492975476033e-07,
+      "loss": 0.1546,
+      "step": 3459
+    },
+    {
+      "epoch": 0.7795645928971752,
+      "grad_norm": 0.48384751710637675,
+      "learning_rate": 6.35008183105704e-07,
+      "loss": 0.1525,
+      "step": 3460
+    },
+    {
+      "epoch": 0.7797899005829836,
+      "grad_norm": 0.4526387079979141,
+      "learning_rate": 6.337681042242447e-07,
+      "loss": 0.1344,
+      "step": 3461
+    },
+    {
+      "epoch": 0.7800152082687921,
+      "grad_norm": 0.4617928977357517,
+      "learning_rate": 6.325290615917961e-07,
+      "loss": 0.1385,
+      "step": 3462
+    },
+    {
+      "epoch": 0.7802405159546005,
+      "grad_norm": 0.5171721464243675,
+      "learning_rate": 6.312910558963505e-07,
+      "loss": 0.1577,
+      "step": 3463
+    },
+    {
+      "epoch": 0.780465823640409,
+      "grad_norm": 0.4784660703129677,
+      "learning_rate": 6.300540878253286e-07,
+      "loss": 0.1531,
+      "step": 3464
+    },
+    {
+      "epoch": 0.7806911313262174,
+      "grad_norm": 0.4828939193700581,
+      "learning_rate": 6.288181580655709e-07,
+      "loss": 0.1401,
+      "step": 3465
+    },
+    {
+      "epoch": 0.7809164390120258,
+      "grad_norm": 0.47170142591052516,
+      "learning_rate": 6.27583267303343e-07,
+      "loss": 0.1431,
+      "step": 3466
+    },
+    {
+      "epoch": 0.7811417466978342,
+      "grad_norm": 0.4754041163728465,
+      "learning_rate": 6.263494162243352e-07,
+      "loss": 0.1525,
+      "step": 3467
+    },
+    {
+      "epoch": 0.7813670543836426,
+      "grad_norm": 0.5184205396388687,
+      "learning_rate": 6.251166055136573e-07,
+      "loss": 0.1569,
+      "step": 3468
+    },
+    {
+      "epoch": 0.781592362069451,
+      "grad_norm": 0.44971934179254847,
+      "learning_rate": 6.238848358558439e-07,
+      "loss": 0.1377,
+      "step": 3469
+    },
+    {
+      "epoch": 0.7818176697552596,
+      "grad_norm": 0.4819646264780796,
+      "learning_rate": 6.226541079348517e-07,
+      "loss": 0.147,
+      "step": 3470
+    },
+    {
+      "epoch": 0.782042977441068,
+      "grad_norm": 0.48149851981924063,
+      "learning_rate": 6.214244224340563e-07,
+      "loss": 0.1371,
+      "step": 3471
+    },
+    {
+      "epoch": 0.7822682851268764,
+      "grad_norm": 0.475720575596046,
+      "learning_rate": 6.201957800362579e-07,
+      "loss": 0.141,
+      "step": 3472
+    },
+    {
+      "epoch": 0.7824935928126848,
+      "grad_norm": 0.47487728061467577,
+      "learning_rate": 6.189681814236742e-07,
+      "loss": 0.1431,
+      "step": 3473
+    },
+    {
+      "epoch": 0.7827189004984932,
+      "grad_norm": 0.48977760168102874,
+      "learning_rate": 6.177416272779468e-07,
+      "loss": 0.1526,
+      "step": 3474
+    },
+    {
+      "epoch": 0.7829442081843017,
+      "grad_norm": 0.4541128559164801,
+      "learning_rate": 6.165161182801336e-07,
+      "loss": 0.1338,
+      "step": 3475
+    },
+    {
+      "epoch": 0.7831695158701101,
+      "grad_norm": 0.4974775276030226,
+      "learning_rate": 6.152916551107149e-07,
+      "loss": 0.1489,
+      "step": 3476
+    },
+    {
+      "epoch": 0.7833948235559185,
+      "grad_norm": 0.47432347761436366,
+      "learning_rate": 6.140682384495902e-07,
+      "loss": 0.1418,
+      "step": 3477
+    },
+    {
+      "epoch": 0.783620131241727,
+      "grad_norm": 0.4693783520677733,
+      "learning_rate": 6.12845868976076e-07,
+      "loss": 0.1478,
+      "step": 3478
+    },
+    {
+      "epoch": 0.7838454389275354,
+      "grad_norm": 0.45579060670234034,
+      "learning_rate": 6.116245473689094e-07,
+      "loss": 0.1404,
+      "step": 3479
+    },
+    {
+      "epoch": 0.7840707466133439,
+      "grad_norm": 0.4825732781016127,
+      "learning_rate": 6.104042743062439e-07,
+      "loss": 0.1557,
+      "step": 3480
+    },
+    {
+      "epoch": 0.7842960542991523,
+      "grad_norm": 0.471553522747966,
+      "learning_rate": 6.091850504656527e-07,
+      "loss": 0.1448,
+      "step": 3481
+    },
+    {
+      "epoch": 0.7845213619849607,
+      "grad_norm": 0.4573226658022487,
+      "learning_rate": 6.079668765241248e-07,
+      "loss": 0.1273,
+      "step": 3482
+    },
+    {
+      "epoch": 0.7847466696707691,
+      "grad_norm": 0.5068814042932546,
+      "learning_rate": 6.06749753158066e-07,
+      "loss": 0.1488,
+      "step": 3483
+    },
+    {
+      "epoch": 0.7849719773565775,
+      "grad_norm": 0.46266260496163186,
+      "learning_rate": 6.05533681043301e-07,
+      "loss": 0.1481,
+      "step": 3484
+    },
+    {
+      "epoch": 0.7851972850423861,
+      "grad_norm": 0.47087969493676185,
+      "learning_rate": 6.04318660855068e-07,
+      "loss": 0.1373,
+      "step": 3485
+    },
+    {
+      "epoch": 0.7854225927281945,
+      "grad_norm": 0.46922056971343673,
+      "learning_rate": 6.031046932680229e-07,
+      "loss": 0.1436,
+      "step": 3486
+    },
+    {
+      "epoch": 0.7856479004140029,
+      "grad_norm": 0.5261567511445522,
+      "learning_rate": 6.018917789562372e-07,
+      "loss": 0.1581,
+      "step": 3487
+    },
+    {
+      "epoch": 0.7858732080998113,
+      "grad_norm": 0.4666125182311981,
+      "learning_rate": 6.006799185931964e-07,
+      "loss": 0.137,
+      "step": 3488
+    },
+    {
+      "epoch": 0.7860985157856197,
+      "grad_norm": 0.48593232146737003,
+      "learning_rate": 5.994691128518019e-07,
+      "loss": 0.1418,
+      "step": 3489
+    },
+    {
+      "epoch": 0.7863238234714282,
+      "grad_norm": 0.5015236294415071,
+      "learning_rate": 5.982593624043682e-07,
+      "loss": 0.1542,
+      "step": 3490
+    },
+    {
+      "epoch": 0.7865491311572366,
+      "grad_norm": 0.47338585550849926,
+      "learning_rate": 5.970506679226249e-07,
+      "loss": 0.1461,
+      "step": 3491
+    },
+    {
+      "epoch": 0.786774438843045,
+      "grad_norm": 0.47921247062245814,
+      "learning_rate": 5.958430300777157e-07,
+      "loss": 0.1438,
+      "step": 3492
+    },
+    {
+      "epoch": 0.7869997465288535,
+      "grad_norm": 0.46779328934505154,
+      "learning_rate": 5.94636449540196e-07,
+      "loss": 0.1439,
+      "step": 3493
+    },
+    {
+      "epoch": 0.7872250542146619,
+      "grad_norm": 0.48007232443930653,
+      "learning_rate": 5.934309269800359e-07,
+      "loss": 0.1523,
+      "step": 3494
+    },
+    {
+      "epoch": 0.7874503619004704,
+      "grad_norm": 0.46819545489036524,
+      "learning_rate": 5.922264630666161e-07,
+      "loss": 0.1396,
+      "step": 3495
+    },
+    {
+      "epoch": 0.7876756695862788,
+      "grad_norm": 0.4786971066876147,
+      "learning_rate": 5.910230584687316e-07,
+      "loss": 0.1465,
+      "step": 3496
+    },
+    {
+      "epoch": 0.7879009772720872,
+      "grad_norm": 0.47295526960505263,
+      "learning_rate": 5.898207138545867e-07,
+      "loss": 0.1441,
+      "step": 3497
+    },
+    {
+      "epoch": 0.7881262849578956,
+      "grad_norm": 0.4953715652067572,
+      "learning_rate": 5.886194298917994e-07,
+      "loss": 0.1446,
+      "step": 3498
+    },
+    {
+      "epoch": 0.788351592643704,
+      "grad_norm": 0.48175106893532155,
+      "learning_rate": 5.874192072473995e-07,
+      "loss": 0.1481,
+      "step": 3499
+    },
+    {
+      "epoch": 0.7885769003295124,
+      "grad_norm": 0.4930876648821759,
+      "learning_rate": 5.862200465878228e-07,
+      "loss": 0.148,
+      "step": 3500
+    },
+    {
+      "epoch": 0.7885769003295124,
+      "eval_loss": 0.14415566623210907,
+      "eval_runtime": 56.9444,
+      "eval_samples_per_second": 50.4,
+      "eval_steps_per_second": 6.304,
+      "step": 3500
+    },
+    {
+      "epoch": 0.788802208015321,
+      "grad_norm": 0.4744664460269437,
+      "learning_rate": 5.850219485789199e-07,
+      "loss": 0.1536,
+      "step": 3501
+    },
+    {
+      "epoch": 0.7890275157011294,
+      "grad_norm": 0.4700558325774627,
+      "learning_rate": 5.838249138859509e-07,
+      "loss": 0.1479,
+      "step": 3502
+    },
+    {
+      "epoch": 0.7892528233869378,
+      "grad_norm": 0.47987032673527097,
+      "learning_rate": 5.826289431735832e-07,
+      "loss": 0.1479,
+      "step": 3503
+    },
+    {
+      "epoch": 0.7894781310727462,
+      "grad_norm": 0.47586202385382154,
+      "learning_rate": 5.814340371058957e-07,
+      "loss": 0.1392,
+      "step": 3504
+    },
+    {
+      "epoch": 0.7897034387585546,
+      "grad_norm": 0.45872762028622155,
+      "learning_rate": 5.802401963463741e-07,
+      "loss": 0.1366,
+      "step": 3505
+    },
+    {
+      "epoch": 0.7899287464443631,
+      "grad_norm": 0.4999088673800423,
+      "learning_rate": 5.79047421557915e-07,
+      "loss": 0.1552,
+      "step": 3506
+    },
+    {
+      "epoch": 0.7901540541301715,
+      "grad_norm": 0.488380827577179,
+      "learning_rate": 5.778557134028207e-07,
+      "loss": 0.1455,
+      "step": 3507
+    },
+    {
+      "epoch": 0.7903793618159799,
+      "grad_norm": 0.4821400153281917,
+      "learning_rate": 5.766650725428027e-07,
+      "loss": 0.1451,
+      "step": 3508
+    },
+    {
+      "epoch": 0.7906046695017884,
+      "grad_norm": 0.44518148929688545,
+      "learning_rate": 5.754754996389799e-07,
+      "loss": 0.1244,
+      "step": 3509
+    },
+    {
+      "epoch": 0.7908299771875968,
+      "grad_norm": 0.49627472177673576,
+      "learning_rate": 5.742869953518773e-07,
+      "loss": 0.1621,
+      "step": 3510
+    },
+    {
+      "epoch": 0.7910552848734053,
+      "grad_norm": 0.4709678650005438,
+      "learning_rate": 5.730995603414274e-07,
+      "loss": 0.1398,
+      "step": 3511
+    },
+    {
+      "epoch": 0.7912805925592137,
+      "grad_norm": 0.4995264521273658,
+      "learning_rate": 5.719131952669679e-07,
+      "loss": 0.1573,
+      "step": 3512
+    },
+    {
+      "epoch": 0.7915059002450221,
+      "grad_norm": 0.4935413382590102,
+      "learning_rate": 5.707279007872435e-07,
+      "loss": 0.1443,
+      "step": 3513
+    },
+    {
+      "epoch": 0.7917312079308305,
+      "grad_norm": 0.46811740322179435,
+      "learning_rate": 5.695436775604049e-07,
+      "loss": 0.1436,
+      "step": 3514
+    },
+    {
+      "epoch": 0.7919565156166389,
+      "grad_norm": 0.47718189498348085,
+      "learning_rate": 5.683605262440056e-07,
+      "loss": 0.1478,
+      "step": 3515
+    },
+    {
+      "epoch": 0.7921818233024474,
+      "grad_norm": 0.4513705321571012,
+      "learning_rate": 5.671784474950068e-07,
+      "loss": 0.1283,
+      "step": 3516
+    },
+    {
+      "epoch": 0.7924071309882559,
+      "grad_norm": 0.48768074675572204,
+      "learning_rate": 5.659974419697723e-07,
+      "loss": 0.1492,
+      "step": 3517
+    },
+    {
+      "epoch": 0.7926324386740643,
+      "grad_norm": 0.47371788564821754,
+      "learning_rate": 5.648175103240694e-07,
+      "loss": 0.142,
+      "step": 3518
+    },
+    {
+      "epoch": 0.7928577463598727,
+      "grad_norm": 0.48134969198220917,
+      "learning_rate": 5.636386532130717e-07,
+      "loss": 0.1554,
+      "step": 3519
+    },
+    {
+      "epoch": 0.7930830540456811,
+      "grad_norm": 0.47825140693570056,
+      "learning_rate": 5.624608712913531e-07,
+      "loss": 0.1392,
+      "step": 3520
+    },
+    {
+      "epoch": 0.7933083617314896,
+      "grad_norm": 0.4958972676890111,
+      "learning_rate": 5.612841652128939e-07,
+      "loss": 0.1534,
+      "step": 3521
+    },
+    {
+      "epoch": 0.793533669417298,
+      "grad_norm": 0.48208591825823033,
+      "learning_rate": 5.601085356310734e-07,
+      "loss": 0.1321,
+      "step": 3522
+    },
+    {
+      "epoch": 0.7937589771031064,
+      "grad_norm": 0.48139189257298015,
+      "learning_rate": 5.589339831986754e-07,
+      "loss": 0.1535,
+      "step": 3523
+    },
+    {
+      "epoch": 0.7939842847889148,
+      "grad_norm": 0.5076369107122871,
+      "learning_rate": 5.577605085678858e-07,
+      "loss": 0.1548,
+      "step": 3524
+    },
+    {
+      "epoch": 0.7942095924747233,
+      "grad_norm": 0.4814164020663355,
+      "learning_rate": 5.565881123902903e-07,
+      "loss": 0.1352,
+      "step": 3525
+    },
+    {
+      "epoch": 0.7944349001605318,
+      "grad_norm": 0.4949034169556831,
+      "learning_rate": 5.554167953168779e-07,
+      "loss": 0.1441,
+      "step": 3526
+    },
+    {
+      "epoch": 0.7946602078463402,
+      "grad_norm": 0.489097629639028,
+      "learning_rate": 5.542465579980361e-07,
+      "loss": 0.1508,
+      "step": 3527
+    },
+    {
+      "epoch": 0.7948855155321486,
+      "grad_norm": 0.4954683894495445,
+      "learning_rate": 5.530774010835552e-07,
+      "loss": 0.1521,
+      "step": 3528
+    },
+    {
+      "epoch": 0.795110823217957,
+      "grad_norm": 0.4558862088107603,
+      "learning_rate": 5.519093252226232e-07,
+      "loss": 0.1352,
+      "step": 3529
+    },
+    {
+      "epoch": 0.7953361309037654,
+      "grad_norm": 0.4767572914325259,
+      "learning_rate": 5.507423310638299e-07,
+      "loss": 0.1439,
+      "step": 3530
+    },
+    {
+      "epoch": 0.7955614385895738,
+      "grad_norm": 0.4933828683243226,
+      "learning_rate": 5.495764192551642e-07,
+      "loss": 0.1428,
+      "step": 3531
+    },
+    {
+      "epoch": 0.7957867462753824,
+      "grad_norm": 0.4790258722048279,
+      "learning_rate": 5.48411590444012e-07,
+      "loss": 0.1498,
+      "step": 3532
+    },
+    {
+      "epoch": 0.7960120539611908,
+      "grad_norm": 0.48235971021995777,
+      "learning_rate": 5.47247845277161e-07,
+      "loss": 0.1502,
+      "step": 3533
+    },
+    {
+      "epoch": 0.7962373616469992,
+      "grad_norm": 0.48860625321354456,
+      "learning_rate": 5.460851844007945e-07,
+      "loss": 0.1597,
+      "step": 3534
+    },
+    {
+      "epoch": 0.7964626693328076,
+      "grad_norm": 0.43448018095407626,
+      "learning_rate": 5.449236084604942e-07,
+      "loss": 0.1199,
+      "step": 3535
+    },
+    {
+      "epoch": 0.796687977018616,
+      "grad_norm": 0.4937221929181501,
+      "learning_rate": 5.437631181012415e-07,
+      "loss": 0.1469,
+      "step": 3536
+    },
+    {
+      "epoch": 0.7969132847044245,
+      "grad_norm": 0.47705394970016807,
+      "learning_rate": 5.426037139674117e-07,
+      "loss": 0.136,
+      "step": 3537
+    },
+    {
+      "epoch": 0.7971385923902329,
+      "grad_norm": 0.4931225755816937,
+      "learning_rate": 5.414453967027797e-07,
+      "loss": 0.1495,
+      "step": 3538
+    },
+    {
+      "epoch": 0.7973639000760413,
+      "grad_norm": 0.5226037448579405,
+      "learning_rate": 5.402881669505164e-07,
+      "loss": 0.16,
+      "step": 3539
+    },
+    {
+      "epoch": 0.7975892077618498,
+      "grad_norm": 0.4760864953251087,
+      "learning_rate": 5.391320253531868e-07,
+      "loss": 0.1394,
+      "step": 3540
+    },
+    {
+      "epoch": 0.7978145154476582,
+      "grad_norm": 0.4896281847697234,
+      "learning_rate": 5.37976972552755e-07,
+      "loss": 0.1512,
+      "step": 3541
+    },
+    {
+      "epoch": 0.7980398231334667,
+      "grad_norm": 0.4619253255008005,
+      "learning_rate": 5.368230091905774e-07,
+      "loss": 0.1391,
+      "step": 3542
+    },
+    {
+      "epoch": 0.7982651308192751,
+      "grad_norm": 0.4886674760091969,
+      "learning_rate": 5.356701359074076e-07,
+      "loss": 0.1428,
+      "step": 3543
+    },
+    {
+      "epoch": 0.7984904385050835,
+      "grad_norm": 0.4714475722790928,
+      "learning_rate": 5.345183533433926e-07,
+      "loss": 0.1422,
+      "step": 3544
+    },
+    {
+      "epoch": 0.7987157461908919,
+      "grad_norm": 0.46688084342141173,
+      "learning_rate": 5.333676621380746e-07,
+      "loss": 0.1326,
+      "step": 3545
+    },
+    {
+      "epoch": 0.7989410538767003,
+      "grad_norm": 0.46759152413970434,
+      "learning_rate": 5.322180629303902e-07,
+      "loss": 0.1349,
+      "step": 3546
+    },
+    {
+      "epoch": 0.7991663615625088,
+      "grad_norm": 0.4718643364907308,
+      "learning_rate": 5.310695563586676e-07,
+      "loss": 0.1363,
+      "step": 3547
+    },
+    {
+      "epoch": 0.7993916692483173,
+      "grad_norm": 0.48857271578568195,
+      "learning_rate": 5.299221430606313e-07,
+      "loss": 0.1488,
+      "step": 3548
+    },
+    {
+      "epoch": 0.7996169769341257,
+      "grad_norm": 0.4940060034052037,
+      "learning_rate": 5.287758236733956e-07,
+      "loss": 0.1427,
+      "step": 3549
+    },
+    {
+      "epoch": 0.7998422846199341,
+      "grad_norm": 0.45653832212422973,
+      "learning_rate": 5.276305988334701e-07,
+      "loss": 0.1312,
+      "step": 3550
+    },
+    {
+      "epoch": 0.8000675923057425,
+      "grad_norm": 0.4807674425195621,
+      "learning_rate": 5.264864691767551e-07,
+      "loss": 0.1359,
+      "step": 3551
+    },
+    {
+      "epoch": 0.800292899991551,
+      "grad_norm": 0.49869397116767594,
+      "learning_rate": 5.253434353385422e-07,
+      "loss": 0.1508,
+      "step": 3552
+    },
+    {
+      "epoch": 0.8005182076773594,
+      "grad_norm": 0.49196778717615613,
+      "learning_rate": 5.242014979535173e-07,
+      "loss": 0.1441,
+      "step": 3553
+    },
+    {
+      "epoch": 0.8007435153631678,
+      "grad_norm": 0.4646584950573559,
+      "learning_rate": 5.23060657655754e-07,
+      "loss": 0.1283,
+      "step": 3554
+    },
+    {
+      "epoch": 0.8009688230489762,
+      "grad_norm": 0.4672535299105578,
+      "learning_rate": 5.219209150787189e-07,
+      "loss": 0.1353,
+      "step": 3555
+    },
+    {
+      "epoch": 0.8011941307347847,
+      "grad_norm": 0.4700008614184278,
+      "learning_rate": 5.207822708552695e-07,
+      "loss": 0.1387,
+      "step": 3556
+    },
+    {
+      "epoch": 0.8014194384205932,
+      "grad_norm": 0.4629867095595067,
+      "learning_rate": 5.196447256176509e-07,
+      "loss": 0.1364,
+      "step": 3557
+    },
+    {
+      "epoch": 0.8016447461064016,
+      "grad_norm": 0.44529648280702133,
+      "learning_rate": 5.185082799975013e-07,
+      "loss": 0.1351,
+      "step": 3558
+    },
+    {
+      "epoch": 0.80187005379221,
+      "grad_norm": 0.494508296543035,
+      "learning_rate": 5.173729346258452e-07,
+      "loss": 0.1463,
+      "step": 3559
+    },
+    {
+      "epoch": 0.8020953614780184,
+      "grad_norm": 0.48079362147959903,
+      "learning_rate": 5.162386901330977e-07,
+      "loss": 0.1515,
+      "step": 3560
+    },
+    {
+      "epoch": 0.8023206691638268,
+      "grad_norm": 0.46475195640142025,
+      "learning_rate": 5.151055471490638e-07,
+      "loss": 0.1438,
+      "step": 3561
+    },
+    {
+      "epoch": 0.8025459768496352,
+      "grad_norm": 0.45803423853802877,
+      "learning_rate": 5.139735063029338e-07,
+      "loss": 0.1306,
+      "step": 3562
+    },
+    {
+      "epoch": 0.8027712845354437,
+      "grad_norm": 0.4891276069352578,
+      "learning_rate": 5.128425682232893e-07,
+      "loss": 0.1444,
+      "step": 3563
+    },
+    {
+      "epoch": 0.8029965922212522,
+      "grad_norm": 0.4673632430066725,
+      "learning_rate": 5.117127335380967e-07,
+      "loss": 0.1419,
+      "step": 3564
+    },
+    {
+      "epoch": 0.8032218999070606,
+      "grad_norm": 0.4530586340062168,
+      "learning_rate": 5.105840028747125e-07,
+      "loss": 0.1365,
+      "step": 3565
+    },
+    {
+      "epoch": 0.803447207592869,
+      "grad_norm": 0.43786476538662233,
+      "learning_rate": 5.094563768598773e-07,
+      "loss": 0.1258,
+      "step": 3566
+    },
+    {
+      "epoch": 0.8036725152786774,
+      "grad_norm": 0.4740710531754334,
+      "learning_rate": 5.083298561197205e-07,
+      "loss": 0.138,
+      "step": 3567
+    },
+    {
+      "epoch": 0.8038978229644859,
+      "grad_norm": 0.45026579308376163,
+      "learning_rate": 5.07204441279758e-07,
+      "loss": 0.1317,
+      "step": 3568
+    },
+    {
+      "epoch": 0.8041231306502943,
+      "grad_norm": 0.4780330236506349,
+      "learning_rate": 5.060801329648896e-07,
+      "loss": 0.1354,
+      "step": 3569
+    },
+    {
+      "epoch": 0.8043484383361027,
+      "grad_norm": 0.4699443411669041,
+      "learning_rate": 5.049569317994013e-07,
+      "loss": 0.1395,
+      "step": 3570
+    },
+    {
+      "epoch": 0.8045737460219111,
+      "grad_norm": 0.47850393654361223,
+      "learning_rate": 5.038348384069663e-07,
+      "loss": 0.1463,
+      "step": 3571
+    },
+    {
+      "epoch": 0.8047990537077196,
+      "grad_norm": 0.4504379700975781,
+      "learning_rate": 5.027138534106399e-07,
+      "loss": 0.1276,
+      "step": 3572
+    },
+    {
+      "epoch": 0.8050243613935281,
+      "grad_norm": 0.4731891026940293,
+      "learning_rate": 5.015939774328643e-07,
+      "loss": 0.1402,
+      "step": 3573
+    },
+    {
+      "epoch": 0.8052496690793365,
+      "grad_norm": 0.4672426623202153,
+      "learning_rate": 5.004752110954642e-07,
+      "loss": 0.1434,
+      "step": 3574
+    },
+    {
+      "epoch": 0.8054749767651449,
+      "grad_norm": 0.4660024619598315,
+      "learning_rate": 4.993575550196495e-07,
+      "loss": 0.1367,
+      "step": 3575
+    },
+    {
+      "epoch": 0.8057002844509533,
+      "grad_norm": 0.5192953141840139,
+      "learning_rate": 4.982410098260118e-07,
+      "loss": 0.164,
+      "step": 3576
+    },
+    {
+      "epoch": 0.8059255921367617,
+      "grad_norm": 0.4927213535021479,
+      "learning_rate": 4.971255761345278e-07,
+      "loss": 0.1457,
+      "step": 3577
+    },
+    {
+      "epoch": 0.8061508998225702,
+      "grad_norm": 0.49996850602429554,
+      "learning_rate": 4.96011254564557e-07,
+      "loss": 0.1545,
+      "step": 3578
+    },
+    {
+      "epoch": 0.8063762075083787,
+      "grad_norm": 0.48978869042351664,
+      "learning_rate": 4.948980457348393e-07,
+      "loss": 0.1507,
+      "step": 3579
+    },
+    {
+      "epoch": 0.8066015151941871,
+      "grad_norm": 0.4945780921266697,
+      "learning_rate": 4.937859502634992e-07,
+      "loss": 0.1531,
+      "step": 3580
+    },
+    {
+      "epoch": 0.8068268228799955,
+      "grad_norm": 0.4926202843506412,
+      "learning_rate": 4.926749687680407e-07,
+      "loss": 0.1445,
+      "step": 3581
+    },
+    {
+      "epoch": 0.8070521305658039,
+      "grad_norm": 0.49038103728173205,
+      "learning_rate": 4.915651018653511e-07,
+      "loss": 0.15,
+      "step": 3582
+    },
+    {
+      "epoch": 0.8072774382516124,
+      "grad_norm": 0.4921243279323083,
+      "learning_rate": 4.904563501716986e-07,
+      "loss": 0.1421,
+      "step": 3583
+    },
+    {
+      "epoch": 0.8075027459374208,
+      "grad_norm": 0.4864573378021889,
+      "learning_rate": 4.893487143027307e-07,
+      "loss": 0.1556,
+      "step": 3584
+    },
+    {
+      "epoch": 0.8077280536232292,
+      "grad_norm": 0.48165400100364,
+      "learning_rate": 4.88242194873477e-07,
+      "loss": 0.1376,
+      "step": 3585
+    },
+    {
+      "epoch": 0.8079533613090376,
+      "grad_norm": 0.46793419647308165,
+      "learning_rate": 4.871367924983458e-07,
+      "loss": 0.1283,
+      "step": 3586
+    },
+    {
+      "epoch": 0.8081786689948461,
+      "grad_norm": 0.4563054622427974,
+      "learning_rate": 4.860325077911271e-07,
+      "loss": 0.1307,
+      "step": 3587
+    },
+    {
+      "epoch": 0.8084039766806546,
+      "grad_norm": 0.4778147728712531,
+      "learning_rate": 4.84929341364988e-07,
+      "loss": 0.1426,
+      "step": 3588
+    },
+    {
+      "epoch": 0.808629284366463,
+      "grad_norm": 0.470678825814214,
+      "learning_rate": 4.838272938324753e-07,
+      "loss": 0.1408,
+      "step": 3589
+    },
+    {
+      "epoch": 0.8088545920522714,
+      "grad_norm": 0.47747949172864607,
+      "learning_rate": 4.827263658055161e-07,
+      "loss": 0.134,
+      "step": 3590
+    },
+    {
+      "epoch": 0.8090798997380798,
+      "grad_norm": 0.45299779005735497,
+      "learning_rate": 4.816265578954135e-07,
+      "loss": 0.1267,
+      "step": 3591
+    },
+    {
+      "epoch": 0.8093052074238882,
+      "grad_norm": 0.47263317790931686,
+      "learning_rate": 4.805278707128505e-07,
+      "loss": 0.1456,
+      "step": 3592
+    },
+    {
+      "epoch": 0.8095305151096966,
+      "grad_norm": 0.46590099895733506,
+      "learning_rate": 4.794303048678878e-07,
+      "loss": 0.1369,
+      "step": 3593
+    },
+    {
+      "epoch": 0.8097558227955051,
+      "grad_norm": 0.4739549078475938,
+      "learning_rate": 4.783338609699614e-07,
+      "loss": 0.1419,
+      "step": 3594
+    },
+    {
+      "epoch": 0.8099811304813136,
+      "grad_norm": 0.482674145715541,
+      "learning_rate": 4.772385396278872e-07,
+      "loss": 0.1426,
+      "step": 3595
+    },
+    {
+      "epoch": 0.810206438167122,
+      "grad_norm": 0.4842047503815621,
+      "learning_rate": 4.7614434144985486e-07,
+      "loss": 0.1493,
+      "step": 3596
+    },
+    {
+      "epoch": 0.8104317458529304,
+      "grad_norm": 0.4688780747440212,
+      "learning_rate": 4.750512670434332e-07,
+      "loss": 0.1442,
+      "step": 3597
+    },
+    {
+      "epoch": 0.8106570535387388,
+      "grad_norm": 0.47451732147636083,
+      "learning_rate": 4.73959317015564e-07,
+      "loss": 0.1415,
+      "step": 3598
+    },
+    {
+      "epoch": 0.8108823612245473,
+      "grad_norm": 0.4738635298914328,
+      "learning_rate": 4.728684919725679e-07,
+      "loss": 0.1395,
+      "step": 3599
+    },
+    {
+      "epoch": 0.8111076689103557,
+      "grad_norm": 0.4828211724052069,
+      "learning_rate": 4.7177879252013945e-07,
+      "loss": 0.1433,
+      "step": 3600
+    },
+    {
+      "epoch": 0.8113329765961641,
+      "grad_norm": 0.46128104439434287,
+      "learning_rate": 4.70690219263347e-07,
+      "loss": 0.1423,
+      "step": 3601
+    },
+    {
+      "epoch": 0.8115582842819725,
+      "grad_norm": 0.4775004231158638,
+      "learning_rate": 4.6960277280663574e-07,
+      "loss": 0.151,
+      "step": 3602
+    },
+    {
+      "epoch": 0.811783591967781,
+      "grad_norm": 0.449064787635663,
+      "learning_rate": 4.685164537538234e-07,
+      "loss": 0.1333,
+      "step": 3603
+    },
+    {
+      "epoch": 0.8120088996535895,
+      "grad_norm": 0.5072565630500627,
+      "learning_rate": 4.674312627081032e-07,
+      "loss": 0.1476,
+      "step": 3604
+    },
+    {
+      "epoch": 0.8122342073393979,
+      "grad_norm": 0.4948390540213146,
+      "learning_rate": 4.6634720027204093e-07,
+      "loss": 0.1487,
+      "step": 3605
+    },
+    {
+      "epoch": 0.8124595150252063,
+      "grad_norm": 0.4747666299747306,
+      "learning_rate": 4.6526426704757545e-07,
+      "loss": 0.1342,
+      "step": 3606
+    },
+    {
+      "epoch": 0.8126848227110147,
+      "grad_norm": 0.45672093470701447,
+      "learning_rate": 4.641824636360195e-07,
+      "loss": 0.1315,
+      "step": 3607
+    },
+    {
+      "epoch": 0.8129101303968231,
+      "grad_norm": 0.4805529418689535,
+      "learning_rate": 4.6310179063805916e-07,
+      "loss": 0.1417,
+      "step": 3608
+    },
+    {
+      "epoch": 0.8131354380826316,
+      "grad_norm": 0.4496617237929147,
+      "learning_rate": 4.620222486537507e-07,
+      "loss": 0.1314,
+      "step": 3609
+    },
+    {
+      "epoch": 0.81336074576844,
+      "grad_norm": 0.4661839377205937,
+      "learning_rate": 4.609438382825246e-07,
+      "loss": 0.1413,
+      "step": 3610
+    },
+    {
+      "epoch": 0.8135860534542485,
+      "grad_norm": 0.4769628903500223,
+      "learning_rate": 4.598665601231805e-07,
+      "loss": 0.1523,
+      "step": 3611
+    },
+    {
+      "epoch": 0.8138113611400569,
+      "grad_norm": 0.4567423398453036,
+      "learning_rate": 4.587904147738925e-07,
+      "loss": 0.146,
+      "step": 3612
+    },
+    {
+      "epoch": 0.8140366688258653,
+      "grad_norm": 0.49509244987764606,
+      "learning_rate": 4.577154028322023e-07,
+      "loss": 0.1484,
+      "step": 3613
+    },
+    {
+      "epoch": 0.8142619765116738,
+      "grad_norm": 0.48460550102516503,
+      "learning_rate": 4.566415248950251e-07,
+      "loss": 0.1357,
+      "step": 3614
+    },
+    {
+      "epoch": 0.8144872841974822,
+      "grad_norm": 0.46882127288324,
+      "learning_rate": 4.555687815586454e-07,
+      "loss": 0.135,
+      "step": 3615
+    },
+    {
+      "epoch": 0.8147125918832906,
+      "grad_norm": 0.4609617391001502,
+      "learning_rate": 4.5449717341871646e-07,
+      "loss": 0.1415,
+      "step": 3616
+    },
+    {
+      "epoch": 0.814937899569099,
+      "grad_norm": 0.45803310017740106,
+      "learning_rate": 4.534267010702639e-07,
+      "loss": 0.1434,
+      "step": 3617
+    },
+    {
+      "epoch": 0.8151632072549074,
+      "grad_norm": 0.47078556309532865,
+      "learning_rate": 4.5235736510767957e-07,
+      "loss": 0.1486,
+      "step": 3618
+    },
+    {
+      "epoch": 0.815388514940716,
+      "grad_norm": 0.4814040730333512,
+      "learning_rate": 4.5128916612472735e-07,
+      "loss": 0.1593,
+      "step": 3619
+    },
+    {
+      "epoch": 0.8156138226265244,
+      "grad_norm": 0.4474212148189157,
+      "learning_rate": 4.5022210471453664e-07,
+      "loss": 0.1288,
+      "step": 3620
+    },
+    {
+      "epoch": 0.8158391303123328,
+      "grad_norm": 0.45554947792372646,
+      "learning_rate": 4.49156181469608e-07,
+      "loss": 0.1327,
+      "step": 3621
+    },
+    {
+      "epoch": 0.8160644379981412,
+      "grad_norm": 0.4692652076043499,
+      "learning_rate": 4.480913969818099e-07,
+      "loss": 0.1431,
+      "step": 3622
+    },
+    {
+      "epoch": 0.8162897456839496,
+      "grad_norm": 0.48211225107706485,
+      "learning_rate": 4.470277518423749e-07,
+      "loss": 0.1391,
+      "step": 3623
+    },
+    {
+      "epoch": 0.816515053369758,
+      "grad_norm": 0.45803483894327357,
+      "learning_rate": 4.4596524664190674e-07,
+      "loss": 0.1372,
+      "step": 3624
+    },
+    {
+      "epoch": 0.8167403610555665,
+      "grad_norm": 0.4596747253969047,
+      "learning_rate": 4.449038819703758e-07,
+      "loss": 0.1334,
+      "step": 3625
+    },
+    {
+      "epoch": 0.816965668741375,
+      "grad_norm": 0.4908321119704428,
+      "learning_rate": 4.4384365841711684e-07,
+      "loss": 0.1503,
+      "step": 3626
+    },
+    {
+      "epoch": 0.8171909764271834,
+      "grad_norm": 0.47937901395744276,
+      "learning_rate": 4.427845765708341e-07,
+      "loss": 0.1486,
+      "step": 3627
+    },
+    {
+      "epoch": 0.8174162841129918,
+      "grad_norm": 0.522368049867057,
+      "learning_rate": 4.417266370195944e-07,
+      "loss": 0.1656,
+      "step": 3628
+    },
+    {
+      "epoch": 0.8176415917988002,
+      "grad_norm": 0.4715115738148847,
+      "learning_rate": 4.406698403508333e-07,
+      "loss": 0.1425,
+      "step": 3629
+    },
+    {
+      "epoch": 0.8178668994846087,
+      "grad_norm": 0.48330696787253935,
+      "learning_rate": 4.3961418715135097e-07,
+      "loss": 0.1375,
+      "step": 3630
+    },
+    {
+      "epoch": 0.8180922071704171,
+      "grad_norm": 0.47495407680447216,
+      "learning_rate": 4.385596780073112e-07,
+      "loss": 0.1405,
+      "step": 3631
+    },
+    {
+      "epoch": 0.8183175148562255,
+      "grad_norm": 0.4605801773738396,
+      "learning_rate": 4.3750631350424456e-07,
+      "loss": 0.1351,
+      "step": 3632
+    },
+    {
+      "epoch": 0.8185428225420339,
+      "grad_norm": 0.44789070717297397,
+      "learning_rate": 4.36454094227044e-07,
+      "loss": 0.1338,
+      "step": 3633
+    },
+    {
+      "epoch": 0.8187681302278424,
+      "grad_norm": 0.45002983584397277,
+      "learning_rate": 4.354030207599691e-07,
+      "loss": 0.1355,
+      "step": 3634
+    },
+    {
+      "epoch": 0.8189934379136509,
+      "grad_norm": 0.4823857101816615,
+      "learning_rate": 4.3435309368664024e-07,
+      "loss": 0.1504,
+      "step": 3635
+    },
+    {
+      "epoch": 0.8192187455994593,
+      "grad_norm": 0.480880910523841,
+      "learning_rate": 4.333043135900436e-07,
+      "loss": 0.1396,
+      "step": 3636
+    },
+    {
+      "epoch": 0.8194440532852677,
+      "grad_norm": 0.4687653665919674,
+      "learning_rate": 4.3225668105252834e-07,
+      "loss": 0.1376,
+      "step": 3637
+    },
+    {
+      "epoch": 0.8196693609710761,
+      "grad_norm": 0.4697261703304461,
+      "learning_rate": 4.312101966558044e-07,
+      "loss": 0.1362,
+      "step": 3638
+    },
+    {
+      "epoch": 0.8198946686568845,
+      "grad_norm": 0.49584678128138465,
+      "learning_rate": 4.3016486098094667e-07,
+      "loss": 0.147,
+      "step": 3639
+    },
+    {
+      "epoch": 0.820119976342693,
+      "grad_norm": 0.4575663797111487,
+      "learning_rate": 4.2912067460839066e-07,
+      "loss": 0.1358,
+      "step": 3640
+    },
+    {
+      "epoch": 0.8203452840285014,
+      "grad_norm": 0.5065216822528078,
+      "learning_rate": 4.280776381179336e-07,
+      "loss": 0.1545,
+      "step": 3641
+    },
+    {
+      "epoch": 0.8205705917143099,
+      "grad_norm": 0.47092622989286775,
+      "learning_rate": 4.2703575208873585e-07,
+      "loss": 0.1347,
+      "step": 3642
+    },
+    {
+      "epoch": 0.8207958994001183,
+      "grad_norm": 0.4660334975585382,
+      "learning_rate": 4.259950170993166e-07,
+      "loss": 0.1434,
+      "step": 3643
+    },
+    {
+      "epoch": 0.8210212070859267,
+      "grad_norm": 0.500680243028102,
+      "learning_rate": 4.2495543372755854e-07,
+      "loss": 0.1444,
+      "step": 3644
+    },
+    {
+      "epoch": 0.8212465147717352,
+      "grad_norm": 0.46330055548885296,
+      "learning_rate": 4.239170025507025e-07,
+      "loss": 0.1367,
+      "step": 3645
+    },
+    {
+      "epoch": 0.8214718224575436,
+      "grad_norm": 0.47460031403910136,
+      "learning_rate": 4.2287972414535084e-07,
+      "loss": 0.1464,
+      "step": 3646
+    },
+    {
+      "epoch": 0.821697130143352,
+      "grad_norm": 0.4730302345220583,
+      "learning_rate": 4.218435990874664e-07,
+      "loss": 0.1364,
+      "step": 3647
+    },
+    {
+      "epoch": 0.8219224378291604,
+      "grad_norm": 0.47555182850863376,
+      "learning_rate": 4.208086279523699e-07,
+      "loss": 0.1409,
+      "step": 3648
+    },
+    {
+      "epoch": 0.8221477455149688,
+      "grad_norm": 0.48364053815683355,
+      "learning_rate": 4.197748113147429e-07,
+      "loss": 0.1538,
+      "step": 3649
+    },
+    {
+      "epoch": 0.8223730532007774,
+      "grad_norm": 0.4946534870674275,
+      "learning_rate": 4.1874214974862436e-07,
+      "loss": 0.152,
+      "step": 3650
+    },
+    {
+      "epoch": 0.8225983608865858,
+      "grad_norm": 0.4978085337800147,
+      "learning_rate": 4.177106438274131e-07,
+      "loss": 0.1457,
+      "step": 3651
+    },
+    {
+      "epoch": 0.8228236685723942,
+      "grad_norm": 0.46412499698557874,
+      "learning_rate": 4.1668029412386677e-07,
+      "loss": 0.1384,
+      "step": 3652
+    },
+    {
+      "epoch": 0.8230489762582026,
+      "grad_norm": 0.48937384930554245,
+      "learning_rate": 4.1565110121009886e-07,
+      "loss": 0.1535,
+      "step": 3653
+    },
+    {
+      "epoch": 0.823274283944011,
+      "grad_norm": 0.5000518434004941,
+      "learning_rate": 4.146230656575831e-07,
+      "loss": 0.1603,
+      "step": 3654
+    },
+    {
+      "epoch": 0.8234995916298194,
+      "grad_norm": 0.5005304792524752,
+      "learning_rate": 4.1359618803714805e-07,
+      "loss": 0.1618,
+      "step": 3655
+    },
+    {
+      "epoch": 0.8237248993156279,
+      "grad_norm": 0.5345403594608715,
+      "learning_rate": 4.125704689189819e-07,
+      "loss": 0.1702,
+      "step": 3656
+    },
+    {
+      "epoch": 0.8239502070014363,
+      "grad_norm": 0.4724524810649259,
+      "learning_rate": 4.115459088726273e-07,
+      "loss": 0.1545,
+      "step": 3657
+    },
+    {
+      "epoch": 0.8241755146872448,
+      "grad_norm": 0.487616351503573,
+      "learning_rate": 4.105225084669839e-07,
+      "loss": 0.1563,
+      "step": 3658
+    },
+    {
+      "epoch": 0.8244008223730532,
+      "grad_norm": 0.5008318927731561,
+      "learning_rate": 4.095002682703092e-07,
+      "loss": 0.1487,
+      "step": 3659
+    },
+    {
+      "epoch": 0.8246261300588617,
+      "grad_norm": 0.46296159400562426,
+      "learning_rate": 4.084791888502135e-07,
+      "loss": 0.1375,
+      "step": 3660
+    },
+    {
+      "epoch": 0.8248514377446701,
+      "grad_norm": 0.46774301209203334,
+      "learning_rate": 4.0745927077366493e-07,
+      "loss": 0.1417,
+      "step": 3661
+    },
+    {
+      "epoch": 0.8250767454304785,
+      "grad_norm": 0.4798585938930743,
+      "learning_rate": 4.0644051460698634e-07,
+      "loss": 0.1445,
+      "step": 3662
+    },
+    {
+      "epoch": 0.8253020531162869,
+      "grad_norm": 0.4636078969381435,
+      "learning_rate": 4.0542292091585447e-07,
+      "loss": 0.1315,
+      "step": 3663
+    },
+    {
+      "epoch": 0.8255273608020953,
+      "grad_norm": 0.47110940624017383,
+      "learning_rate": 4.0440649026530166e-07,
+      "loss": 0.1437,
+      "step": 3664
+    },
+    {
+      "epoch": 0.8257526684879037,
+      "grad_norm": 0.4845380348187428,
+      "learning_rate": 4.033912232197132e-07,
+      "loss": 0.1471,
+      "step": 3665
+    },
+    {
+      "epoch": 0.8259779761737123,
+      "grad_norm": 0.49698390602523673,
+      "learning_rate": 4.0237712034283004e-07,
+      "loss": 0.1339,
+      "step": 3666
+    },
+    {
+      "epoch": 0.8262032838595207,
+      "grad_norm": 0.4709200362568471,
+      "learning_rate": 4.0136418219774457e-07,
+      "loss": 0.1363,
+      "step": 3667
+    },
+    {
+      "epoch": 0.8264285915453291,
+      "grad_norm": 0.47535675030537516,
+      "learning_rate": 4.003524093469041e-07,
+      "loss": 0.1335,
+      "step": 3668
+    },
+    {
+      "epoch": 0.8266538992311375,
+      "grad_norm": 0.4954697566054315,
+      "learning_rate": 3.993418023521092e-07,
+      "loss": 0.1519,
+      "step": 3669
+    },
+    {
+      "epoch": 0.8268792069169459,
+      "grad_norm": 0.45705098958634477,
+      "learning_rate": 3.983323617745111e-07,
+      "loss": 0.1451,
+      "step": 3670
+    },
+    {
+      "epoch": 0.8271045146027544,
+      "grad_norm": 0.4772107912339602,
+      "learning_rate": 3.9732408817461544e-07,
+      "loss": 0.1357,
+      "step": 3671
+    },
+    {
+      "epoch": 0.8273298222885628,
+      "grad_norm": 0.4540939185349331,
+      "learning_rate": 3.963169821122778e-07,
+      "loss": 0.1238,
+      "step": 3672
+    },
+    {
+      "epoch": 0.8275551299743712,
+      "grad_norm": 0.47655176367622293,
+      "learning_rate": 3.953110441467073e-07,
+      "loss": 0.1295,
+      "step": 3673
+    },
+    {
+      "epoch": 0.8277804376601797,
+      "grad_norm": 0.4974421135266379,
+      "learning_rate": 3.943062748364651e-07,
+      "loss": 0.1524,
+      "step": 3674
+    },
+    {
+      "epoch": 0.8280057453459881,
+      "grad_norm": 0.4630936824471273,
+      "learning_rate": 3.9330267473945973e-07,
+      "loss": 0.1373,
+      "step": 3675
+    },
+    {
+      "epoch": 0.8282310530317966,
+      "grad_norm": 0.47957905381516513,
+      "learning_rate": 3.9230024441295394e-07,
+      "loss": 0.1481,
+      "step": 3676
+    },
+    {
+      "epoch": 0.828456360717605,
+      "grad_norm": 0.47595799930886623,
+      "learning_rate": 3.9129898441356064e-07,
+      "loss": 0.1417,
+      "step": 3677
+    },
+    {
+      "epoch": 0.8286816684034134,
+      "grad_norm": 0.4795537018708493,
+      "learning_rate": 3.9029889529724113e-07,
+      "loss": 0.1355,
+      "step": 3678
+    },
+    {
+      "epoch": 0.8289069760892218,
+      "grad_norm": 0.4767115392349057,
+      "learning_rate": 3.892999776193085e-07,
+      "loss": 0.1412,
+      "step": 3679
+    },
+    {
+      "epoch": 0.8291322837750302,
+      "grad_norm": 0.45880264666880477,
+      "learning_rate": 3.8830223193442345e-07,
+      "loss": 0.126,
+      "step": 3680
+    },
+    {
+      "epoch": 0.8293575914608388,
+      "grad_norm": 0.4849367733903804,
+      "learning_rate": 3.8730565879659845e-07,
+      "loss": 0.1536,
+      "step": 3681
+    },
+    {
+      "epoch": 0.8295828991466472,
+      "grad_norm": 0.4679528634230345,
+      "learning_rate": 3.863102587591919e-07,
+      "loss": 0.1395,
+      "step": 3682
+    },
+    {
+      "epoch": 0.8298082068324556,
+      "grad_norm": 0.5266202362340566,
+      "learning_rate": 3.853160323749128e-07,
+      "loss": 0.1748,
+      "step": 3683
+    },
+    {
+      "epoch": 0.830033514518264,
+      "grad_norm": 0.5021640642613251,
+      "learning_rate": 3.84322980195819e-07,
+      "loss": 0.1464,
+      "step": 3684
+    },
+    {
+      "epoch": 0.8302588222040724,
+      "grad_norm": 0.5141221995124274,
+      "learning_rate": 3.833311027733139e-07,
+      "loss": 0.1631,
+      "step": 3685
+    },
+    {
+      "epoch": 0.8304841298898809,
+      "grad_norm": 0.4973624549310441,
+      "learning_rate": 3.823404006581513e-07,
+      "loss": 0.1418,
+      "step": 3686
+    },
+    {
+      "epoch": 0.8307094375756893,
+      "grad_norm": 0.47186531333299525,
+      "learning_rate": 3.8135087440043017e-07,
+      "loss": 0.1404,
+      "step": 3687
+    },
+    {
+      "epoch": 0.8309347452614977,
+      "grad_norm": 0.4951193897109337,
+      "learning_rate": 3.8036252454959844e-07,
+      "loss": 0.1584,
+      "step": 3688
+    },
+    {
+      "epoch": 0.8311600529473062,
+      "grad_norm": 0.48335252169394144,
+      "learning_rate": 3.7937535165444875e-07,
+      "loss": 0.1518,
+      "step": 3689
+    },
+    {
+      "epoch": 0.8313853606331146,
+      "grad_norm": 0.43298312754915047,
+      "learning_rate": 3.7838935626312246e-07,
+      "loss": 0.1217,
+      "step": 3690
+    },
+    {
+      "epoch": 0.831610668318923,
+      "grad_norm": 0.4772264241587924,
+      "learning_rate": 3.7740453892310596e-07,
+      "loss": 0.148,
+      "step": 3691
+    },
+    {
+      "epoch": 0.8318359760047315,
+      "grad_norm": 0.4986281566598413,
+      "learning_rate": 3.764209001812316e-07,
+      "loss": 0.1563,
+      "step": 3692
+    },
+    {
+      "epoch": 0.8320612836905399,
+      "grad_norm": 0.49596555246639523,
+      "learning_rate": 3.754384405836767e-07,
+      "loss": 0.1472,
+      "step": 3693
+    },
+    {
+      "epoch": 0.8322865913763483,
+      "grad_norm": 0.4625550974162768,
+      "learning_rate": 3.7445716067596506e-07,
+      "loss": 0.1367,
+      "step": 3694
+    },
+    {
+      "epoch": 0.8325118990621567,
+      "grad_norm": 0.4707691697369795,
+      "learning_rate": 3.734770610029642e-07,
+      "loss": 0.1478,
+      "step": 3695
+    },
+    {
+      "epoch": 0.8327372067479651,
+      "grad_norm": 0.4863393786315704,
+      "learning_rate": 3.72498142108888e-07,
+      "loss": 0.1467,
+      "step": 3696
+    },
+    {
+      "epoch": 0.8329625144337737,
+      "grad_norm": 0.46568031143045824,
+      "learning_rate": 3.7152040453729223e-07,
+      "loss": 0.1514,
+      "step": 3697
+    },
+    {
+      "epoch": 0.8331878221195821,
+      "grad_norm": 0.4451139600576869,
+      "learning_rate": 3.705438488310792e-07,
+      "loss": 0.1331,
+      "step": 3698
+    },
+    {
+      "epoch": 0.8334131298053905,
+      "grad_norm": 0.4386885790715076,
+      "learning_rate": 3.695684755324938e-07,
+      "loss": 0.1318,
+      "step": 3699
+    },
+    {
+      "epoch": 0.8336384374911989,
+      "grad_norm": 0.4936072561997163,
+      "learning_rate": 3.6859428518312394e-07,
+      "loss": 0.1526,
+      "step": 3700
+    },
+    {
+      "epoch": 0.8338637451770073,
+      "grad_norm": 0.4777576391614967,
+      "learning_rate": 3.6762127832390194e-07,
+      "loss": 0.1527,
+      "step": 3701
+    },
+    {
+      "epoch": 0.8340890528628158,
+      "grad_norm": 0.47375366285599557,
+      "learning_rate": 3.666494554951014e-07,
+      "loss": 0.1347,
+      "step": 3702
+    },
+    {
+      "epoch": 0.8343143605486242,
+      "grad_norm": 0.4708439662424002,
+      "learning_rate": 3.656788172363401e-07,
+      "loss": 0.1276,
+      "step": 3703
+    },
+    {
+      "epoch": 0.8345396682344326,
+      "grad_norm": 0.4710695568080402,
+      "learning_rate": 3.6470936408657647e-07,
+      "loss": 0.144,
+      "step": 3704
+    },
+    {
+      "epoch": 0.8347649759202411,
+      "grad_norm": 0.4974200303377148,
+      "learning_rate": 3.6374109658411207e-07,
+      "loss": 0.1502,
+      "step": 3705
+    },
+    {
+      "epoch": 0.8349902836060495,
+      "grad_norm": 0.4761356662500996,
+      "learning_rate": 3.6277401526659067e-07,
+      "loss": 0.1387,
+      "step": 3706
+    },
+    {
+      "epoch": 0.835215591291858,
+      "grad_norm": 0.4830728913581897,
+      "learning_rate": 3.6180812067099477e-07,
+      "loss": 0.1454,
+      "step": 3707
+    },
+    {
+      "epoch": 0.8354408989776664,
+      "grad_norm": 0.46440622198669923,
+      "learning_rate": 3.6084341333365135e-07,
+      "loss": 0.1389,
+      "step": 3708
+    },
+    {
+      "epoch": 0.8356662066634748,
+      "grad_norm": 0.4782555940658288,
+      "learning_rate": 3.5987989379022536e-07,
+      "loss": 0.1428,
+      "step": 3709
+    },
+    {
+      "epoch": 0.8358915143492832,
+      "grad_norm": 0.44845896618127046,
+      "learning_rate": 3.58917562575723e-07,
+      "loss": 0.1274,
+      "step": 3710
+    },
+    {
+      "epoch": 0.8361168220350916,
+      "grad_norm": 0.4747354070350136,
+      "learning_rate": 3.57956420224492e-07,
+      "loss": 0.1414,
+      "step": 3711
+    },
+    {
+      "epoch": 0.8363421297209,
+      "grad_norm": 0.4788958507356607,
+      "learning_rate": 3.569964672702178e-07,
+      "loss": 0.1409,
+      "step": 3712
+    },
+    {
+      "epoch": 0.8365674374067086,
+      "grad_norm": 0.4583045970863074,
+      "learning_rate": 3.5603770424592785e-07,
+      "loss": 0.1311,
+      "step": 3713
+    },
+    {
+      "epoch": 0.836792745092517,
+      "grad_norm": 0.4714639173787331,
+      "learning_rate": 3.550801316839858e-07,
+      "loss": 0.1331,
+      "step": 3714
+    },
+    {
+      "epoch": 0.8370180527783254,
+      "grad_norm": 0.5013819286501238,
+      "learning_rate": 3.5412375011609714e-07,
+      "loss": 0.1605,
+      "step": 3715
+    },
+    {
+      "epoch": 0.8372433604641338,
+      "grad_norm": 0.46978437287956937,
+      "learning_rate": 3.531685600733051e-07,
+      "loss": 0.1359,
+      "step": 3716
+    },
+    {
+      "epoch": 0.8374686681499423,
+      "grad_norm": 0.4805452566101585,
+      "learning_rate": 3.5221456208598987e-07,
+      "loss": 0.1385,
+      "step": 3717
+    },
+    {
+      "epoch": 0.8376939758357507,
+      "grad_norm": 0.47136760789390303,
+      "learning_rate": 3.5126175668387275e-07,
+      "loss": 0.1421,
+      "step": 3718
+    },
+    {
+      "epoch": 0.8379192835215591,
+      "grad_norm": 0.4763275734278578,
+      "learning_rate": 3.503101443960094e-07,
+      "loss": 0.1445,
+      "step": 3719
+    },
+    {
+      "epoch": 0.8381445912073675,
+      "grad_norm": 0.4838455228665239,
+      "learning_rate": 3.4935972575079524e-07,
+      "loss": 0.1422,
+      "step": 3720
+    },
+    {
+      "epoch": 0.838369898893176,
+      "grad_norm": 0.4592355710517882,
+      "learning_rate": 3.484105012759631e-07,
+      "loss": 0.1359,
+      "step": 3721
+    },
+    {
+      "epoch": 0.8385952065789845,
+      "grad_norm": 0.5076200514758693,
+      "learning_rate": 3.474624714985805e-07,
+      "loss": 0.1509,
+      "step": 3722
+    },
+    {
+      "epoch": 0.8388205142647929,
+      "grad_norm": 0.5013287029320084,
+      "learning_rate": 3.465156369450545e-07,
+      "loss": 0.1507,
+      "step": 3723
+    },
+    {
+      "epoch": 0.8390458219506013,
+      "grad_norm": 0.46463610064480193,
+      "learning_rate": 3.455699981411259e-07,
+      "loss": 0.1386,
+      "step": 3724
+    },
+    {
+      "epoch": 0.8392711296364097,
+      "grad_norm": 0.469792875927176,
+      "learning_rate": 3.446255556118736e-07,
+      "loss": 0.1393,
+      "step": 3725
+    },
+    {
+      "epoch": 0.8394964373222181,
+      "grad_norm": 0.4668736724148676,
+      "learning_rate": 3.436823098817102e-07,
+      "loss": 0.1401,
+      "step": 3726
+    },
+    {
+      "epoch": 0.8397217450080265,
+      "grad_norm": 0.45187758940267403,
+      "learning_rate": 3.427402614743863e-07,
+      "loss": 0.1278,
+      "step": 3727
+    },
+    {
+      "epoch": 0.8399470526938351,
+      "grad_norm": 0.4841475700937596,
+      "learning_rate": 3.417994109129852e-07,
+      "loss": 0.1422,
+      "step": 3728
+    },
+    {
+      "epoch": 0.8401723603796435,
+      "grad_norm": 0.4625871616784283,
+      "learning_rate": 3.408597587199261e-07,
+      "loss": 0.1338,
+      "step": 3729
+    },
+    {
+      "epoch": 0.8403976680654519,
+      "grad_norm": 0.47550866799296265,
+      "learning_rate": 3.3992130541696336e-07,
+      "loss": 0.1369,
+      "step": 3730
+    },
+    {
+      "epoch": 0.8406229757512603,
+      "grad_norm": 0.4490907761475444,
+      "learning_rate": 3.389840515251855e-07,
+      "loss": 0.1339,
+      "step": 3731
+    },
+    {
+      "epoch": 0.8408482834370687,
+      "grad_norm": 0.5026847491625468,
+      "learning_rate": 3.3804799756501335e-07,
+      "loss": 0.1373,
+      "step": 3732
+    },
+    {
+      "epoch": 0.8410735911228772,
+      "grad_norm": 0.4796871566183159,
+      "learning_rate": 3.371131440562042e-07,
+      "loss": 0.1359,
+      "step": 3733
+    },
+    {
+      "epoch": 0.8412988988086856,
+      "grad_norm": 0.489908655988409,
+      "learning_rate": 3.3617949151784623e-07,
+      "loss": 0.1446,
+      "step": 3734
+    },
+    {
+      "epoch": 0.841524206494494,
+      "grad_norm": 0.4611673387819876,
+      "learning_rate": 3.3524704046836305e-07,
+      "loss": 0.146,
+      "step": 3735
+    },
+    {
+      "epoch": 0.8417495141803025,
+      "grad_norm": 0.47569199775126225,
+      "learning_rate": 3.343157914255085e-07,
+      "loss": 0.1313,
+      "step": 3736
+    },
+    {
+      "epoch": 0.841974821866111,
+      "grad_norm": 0.4447117260347802,
+      "learning_rate": 3.3338574490637154e-07,
+      "loss": 0.1307,
+      "step": 3737
+    },
+    {
+      "epoch": 0.8422001295519194,
+      "grad_norm": 0.4541365818430463,
+      "learning_rate": 3.3245690142737236e-07,
+      "loss": 0.1264,
+      "step": 3738
+    },
+    {
+      "epoch": 0.8424254372377278,
+      "grad_norm": 0.4896570773348926,
+      "learning_rate": 3.3152926150426256e-07,
+      "loss": 0.1383,
+      "step": 3739
+    },
+    {
+      "epoch": 0.8426507449235362,
+      "grad_norm": 0.47351248738261037,
+      "learning_rate": 3.306028256521265e-07,
+      "loss": 0.1378,
+      "step": 3740
+    },
+    {
+      "epoch": 0.8428760526093446,
+      "grad_norm": 0.498959457382647,
+      "learning_rate": 3.296775943853789e-07,
+      "loss": 0.1525,
+      "step": 3741
+    },
+    {
+      "epoch": 0.843101360295153,
+      "grad_norm": 0.4913202107412484,
+      "learning_rate": 3.287535682177667e-07,
+      "loss": 0.15,
+      "step": 3742
+    },
+    {
+      "epoch": 0.8433266679809615,
+      "grad_norm": 0.4894838756998072,
+      "learning_rate": 3.278307476623674e-07,
+      "loss": 0.1388,
+      "step": 3743
+    },
+    {
+      "epoch": 0.84355197566677,
+      "grad_norm": 0.4986603478893013,
+      "learning_rate": 3.2690913323158795e-07,
+      "loss": 0.148,
+      "step": 3744
+    },
+    {
+      "epoch": 0.8437772833525784,
+      "grad_norm": 0.5147315679695823,
+      "learning_rate": 3.259887254371677e-07,
+      "loss": 0.1518,
+      "step": 3745
+    },
+    {
+      "epoch": 0.8440025910383868,
+      "grad_norm": 0.5093748124342478,
+      "learning_rate": 3.2506952479017417e-07,
+      "loss": 0.1495,
+      "step": 3746
+    },
+    {
+      "epoch": 0.8442278987241952,
+      "grad_norm": 0.47134348911112495,
+      "learning_rate": 3.241515318010044e-07,
+      "loss": 0.1479,
+      "step": 3747
+    },
+    {
+      "epoch": 0.8444532064100037,
+      "grad_norm": 0.48836861657128516,
+      "learning_rate": 3.2323474697938727e-07,
+      "loss": 0.1525,
+      "step": 3748
+    },
+    {
+      "epoch": 0.8446785140958121,
+      "grad_norm": 0.5219802457338291,
+      "learning_rate": 3.223191708343776e-07,
+      "loss": 0.1558,
+      "step": 3749
+    },
+    {
+      "epoch": 0.8449038217816205,
+      "grad_norm": 0.4751955627178434,
+      "learning_rate": 3.214048038743622e-07,
+      "loss": 0.1405,
+      "step": 3750
+    },
+    {
+      "epoch": 0.8451291294674289,
+      "grad_norm": 0.4602528146172868,
+      "learning_rate": 3.204916466070537e-07,
+      "loss": 0.1369,
+      "step": 3751
+    },
+    {
+      "epoch": 0.8453544371532374,
+      "grad_norm": 0.4867880285193427,
+      "learning_rate": 3.1957969953949506e-07,
+      "loss": 0.1454,
+      "step": 3752
+    },
+    {
+      "epoch": 0.8455797448390459,
+      "grad_norm": 0.47421301891609136,
+      "learning_rate": 3.186689631780565e-07,
+      "loss": 0.139,
+      "step": 3753
+    },
+    {
+      "epoch": 0.8458050525248543,
+      "grad_norm": 0.5230980631330083,
+      "learning_rate": 3.1775943802843546e-07,
+      "loss": 0.1655,
+      "step": 3754
+    },
+    {
+      "epoch": 0.8460303602106627,
+      "grad_norm": 0.4792049371459631,
+      "learning_rate": 3.168511245956582e-07,
+      "loss": 0.1456,
+      "step": 3755
+    },
+    {
+      "epoch": 0.8462556678964711,
+      "grad_norm": 0.4911553003467961,
+      "learning_rate": 3.1594402338407633e-07,
+      "loss": 0.1471,
+      "step": 3756
+    },
+    {
+      "epoch": 0.8464809755822795,
+      "grad_norm": 0.46160048181105984,
+      "learning_rate": 3.1503813489736995e-07,
+      "loss": 0.1351,
+      "step": 3757
+    },
+    {
+      "epoch": 0.846706283268088,
+      "grad_norm": 0.4542443875832939,
+      "learning_rate": 3.141334596385448e-07,
+      "loss": 0.1399,
+      "step": 3758
+    },
+    {
+      "epoch": 0.8469315909538964,
+      "grad_norm": 0.5083072602650772,
+      "learning_rate": 3.132299981099335e-07,
+      "loss": 0.1545,
+      "step": 3759
+    },
+    {
+      "epoch": 0.8471568986397049,
+      "grad_norm": 0.4967217063074372,
+      "learning_rate": 3.12327750813195e-07,
+      "loss": 0.1526,
+      "step": 3760
+    },
+    {
+      "epoch": 0.8473822063255133,
+      "grad_norm": 0.4600415740247211,
+      "learning_rate": 3.1142671824931275e-07,
+      "loss": 0.1382,
+      "step": 3761
+    },
+    {
+      "epoch": 0.8476075140113217,
+      "grad_norm": 0.4800819510127324,
+      "learning_rate": 3.105269009185974e-07,
+      "loss": 0.148,
+      "step": 3762
+    },
+    {
+      "epoch": 0.8478328216971301,
+      "grad_norm": 0.4779608519409477,
+      "learning_rate": 3.096282993206837e-07,
+      "loss": 0.1432,
+      "step": 3763
+    },
+    {
+      "epoch": 0.8480581293829386,
+      "grad_norm": 0.4734713665797522,
+      "learning_rate": 3.087309139545311e-07,
+      "loss": 0.1423,
+      "step": 3764
+    },
+    {
+      "epoch": 0.848283437068747,
+      "grad_norm": 0.4706362568362203,
+      "learning_rate": 3.0783474531842497e-07,
+      "loss": 0.1381,
+      "step": 3765
+    },
+    {
+      "epoch": 0.8485087447545554,
+      "grad_norm": 0.4533526963952121,
+      "learning_rate": 3.0693979390997333e-07,
+      "loss": 0.1279,
+      "step": 3766
+    },
+    {
+      "epoch": 0.8487340524403638,
+      "grad_norm": 0.48340443376537645,
+      "learning_rate": 3.0604606022611033e-07,
+      "loss": 0.1491,
+      "step": 3767
+    },
+    {
+      "epoch": 0.8489593601261723,
+      "grad_norm": 0.4694014430810065,
+      "learning_rate": 3.0515354476309293e-07,
+      "loss": 0.1415,
+      "step": 3768
+    },
+    {
+      "epoch": 0.8491846678119808,
+      "grad_norm": 0.45863277428008126,
+      "learning_rate": 3.042622480165011e-07,
+      "loss": 0.1303,
+      "step": 3769
+    },
+    {
+      "epoch": 0.8494099754977892,
+      "grad_norm": 0.5038764071443956,
+      "learning_rate": 3.033721704812395e-07,
+      "loss": 0.1557,
+      "step": 3770
+    },
+    {
+      "epoch": 0.8496352831835976,
+      "grad_norm": 0.4674809016759538,
+      "learning_rate": 3.024833126515339e-07,
+      "loss": 0.1398,
+      "step": 3771
+    },
+    {
+      "epoch": 0.849860590869406,
+      "grad_norm": 0.5036409352245586,
+      "learning_rate": 3.0159567502093535e-07,
+      "loss": 0.141,
+      "step": 3772
+    },
+    {
+      "epoch": 0.8500858985552144,
+      "grad_norm": 0.46259561465234855,
+      "learning_rate": 3.0070925808231456e-07,
+      "loss": 0.1329,
+      "step": 3773
+    },
+    {
+      "epoch": 0.8503112062410229,
+      "grad_norm": 0.4410361643206542,
+      "learning_rate": 2.9982406232786614e-07,
+      "loss": 0.1299,
+      "step": 3774
+    },
+    {
+      "epoch": 0.8505365139268314,
+      "grad_norm": 0.4491162934123659,
+      "learning_rate": 2.9894008824910726e-07,
+      "loss": 0.1248,
+      "step": 3775
+    },
+    {
+      "epoch": 0.8507618216126398,
+      "grad_norm": 0.4925158953606324,
+      "learning_rate": 2.9805733633687467e-07,
+      "loss": 0.1452,
+      "step": 3776
+    },
+    {
+      "epoch": 0.8509871292984482,
+      "grad_norm": 0.4924716105067092,
+      "learning_rate": 2.9717580708132856e-07,
+      "loss": 0.1406,
+      "step": 3777
+    },
+    {
+      "epoch": 0.8512124369842566,
+      "grad_norm": 0.46442448447501566,
+      "learning_rate": 2.9629550097194787e-07,
+      "loss": 0.1387,
+      "step": 3778
+    },
+    {
+      "epoch": 0.851437744670065,
+      "grad_norm": 0.46692915113918126,
+      "learning_rate": 2.9541641849753557e-07,
+      "loss": 0.14,
+      "step": 3779
+    },
+    {
+      "epoch": 0.8516630523558735,
+      "grad_norm": 0.47585439999813806,
+      "learning_rate": 2.9453856014621224e-07,
+      "loss": 0.1346,
+      "step": 3780
+    },
+    {
+      "epoch": 0.8518883600416819,
+      "grad_norm": 0.45401892300040164,
+      "learning_rate": 2.936619264054194e-07,
+      "loss": 0.1332,
+      "step": 3781
+    },
+    {
+      "epoch": 0.8521136677274903,
+      "grad_norm": 0.46019664304026237,
+      "learning_rate": 2.9278651776192073e-07,
+      "loss": 0.1377,
+      "step": 3782
+    },
+    {
+      "epoch": 0.8523389754132988,
+      "grad_norm": 0.48199717160160416,
+      "learning_rate": 2.919123347017963e-07,
+      "loss": 0.1443,
+      "step": 3783
+    },
+    {
+      "epoch": 0.8525642830991073,
+      "grad_norm": 0.4804863037487851,
+      "learning_rate": 2.910393777104481e-07,
+      "loss": 0.1354,
+      "step": 3784
+    },
+    {
+      "epoch": 0.8527895907849157,
+      "grad_norm": 0.4615664259895404,
+      "learning_rate": 2.901676472725973e-07,
+      "loss": 0.1365,
+      "step": 3785
+    },
+    {
+      "epoch": 0.8530148984707241,
+      "grad_norm": 0.47659295388773687,
+      "learning_rate": 2.892971438722822e-07,
+      "loss": 0.1412,
+      "step": 3786
+    },
+    {
+      "epoch": 0.8532402061565325,
+      "grad_norm": 0.4756320606626126,
+      "learning_rate": 2.8842786799286204e-07,
+      "loss": 0.1454,
+      "step": 3787
+    },
+    {
+      "epoch": 0.8534655138423409,
+      "grad_norm": 0.4668571095658293,
+      "learning_rate": 2.8755982011701183e-07,
+      "loss": 0.1409,
+      "step": 3788
+    },
+    {
+      "epoch": 0.8536908215281493,
+      "grad_norm": 0.4596291324992224,
+      "learning_rate": 2.866930007267274e-07,
+      "loss": 0.1342,
+      "step": 3789
+    },
+    {
+      "epoch": 0.8539161292139578,
+      "grad_norm": 0.4722187049851435,
+      "learning_rate": 2.8582741030332095e-07,
+      "loss": 0.1406,
+      "step": 3790
+    },
+    {
+      "epoch": 0.8541414368997663,
+      "grad_norm": 0.45961773897137925,
+      "learning_rate": 2.8496304932742247e-07,
+      "loss": 0.1355,
+      "step": 3791
+    },
+    {
+      "epoch": 0.8543667445855747,
+      "grad_norm": 0.48484033285073747,
+      "learning_rate": 2.840999182789797e-07,
+      "loss": 0.1407,
+      "step": 3792
+    },
+    {
+      "epoch": 0.8545920522713831,
+      "grad_norm": 0.46030708547491617,
+      "learning_rate": 2.8323801763725623e-07,
+      "loss": 0.1358,
+      "step": 3793
+    },
+    {
+      "epoch": 0.8548173599571915,
+      "grad_norm": 0.5098699348994705,
+      "learning_rate": 2.823773478808348e-07,
+      "loss": 0.1577,
+      "step": 3794
+    },
+    {
+      "epoch": 0.855042667643,
+      "grad_norm": 0.4532517106481429,
+      "learning_rate": 2.8151790948761165e-07,
+      "loss": 0.1349,
+      "step": 3795
+    },
+    {
+      "epoch": 0.8552679753288084,
+      "grad_norm": 0.47392199223502923,
+      "learning_rate": 2.806597029348018e-07,
+      "loss": 0.1495,
+      "step": 3796
+    },
+    {
+      "epoch": 0.8554932830146168,
+      "grad_norm": 0.4720938077032227,
+      "learning_rate": 2.7980272869893633e-07,
+      "loss": 0.1395,
+      "step": 3797
+    },
+    {
+      "epoch": 0.8557185907004252,
+      "grad_norm": 0.48573209575016485,
+      "learning_rate": 2.7894698725585866e-07,
+      "loss": 0.1419,
+      "step": 3798
+    },
+    {
+      "epoch": 0.8559438983862337,
+      "grad_norm": 0.49549561049525453,
+      "learning_rate": 2.7809247908073184e-07,
+      "loss": 0.1599,
+      "step": 3799
+    },
+    {
+      "epoch": 0.8561692060720422,
+      "grad_norm": 0.45867812854518997,
+      "learning_rate": 2.772392046480324e-07,
+      "loss": 0.1354,
+      "step": 3800
+    },
+    {
+      "epoch": 0.8563945137578506,
+      "grad_norm": 0.48750250388740535,
+      "learning_rate": 2.763871644315508e-07,
+      "loss": 0.1453,
+      "step": 3801
+    },
+    {
+      "epoch": 0.856619821443659,
+      "grad_norm": 0.46278874013656607,
+      "learning_rate": 2.755363589043944e-07,
+      "loss": 0.1421,
+      "step": 3802
+    },
+    {
+      "epoch": 0.8568451291294674,
+      "grad_norm": 0.4861918783813127,
+      "learning_rate": 2.746867885389828e-07,
+      "loss": 0.1466,
+      "step": 3803
+    },
+    {
+      "epoch": 0.8570704368152758,
+      "grad_norm": 0.4625862382559484,
+      "learning_rate": 2.738384538070518e-07,
+      "loss": 0.1323,
+      "step": 3804
+    },
+    {
+      "epoch": 0.8572957445010843,
+      "grad_norm": 0.4932943365028231,
+      "learning_rate": 2.7299135517964897e-07,
+      "loss": 0.1514,
+      "step": 3805
+    },
+    {
+      "epoch": 0.8575210521868927,
+      "grad_norm": 0.45459505703034975,
+      "learning_rate": 2.7214549312713723e-07,
+      "loss": 0.1329,
+      "step": 3806
+    },
+    {
+      "epoch": 0.8577463598727012,
+      "grad_norm": 0.4657582875813703,
+      "learning_rate": 2.713008681191923e-07,
+      "loss": 0.1454,
+      "step": 3807
+    },
+    {
+      "epoch": 0.8579716675585096,
+      "grad_norm": 0.44863377479765015,
+      "learning_rate": 2.7045748062480254e-07,
+      "loss": 0.1315,
+      "step": 3808
+    },
+    {
+      "epoch": 0.858196975244318,
+      "grad_norm": 0.4600757566899341,
+      "learning_rate": 2.696153311122704e-07,
+      "loss": 0.1355,
+      "step": 3809
+    },
+    {
+      "epoch": 0.8584222829301265,
+      "grad_norm": 0.4783882851837259,
+      "learning_rate": 2.6877442004920873e-07,
+      "loss": 0.1429,
+      "step": 3810
+    },
+    {
+      "epoch": 0.8586475906159349,
+      "grad_norm": 0.47769058593487773,
+      "learning_rate": 2.6793474790254516e-07,
+      "loss": 0.1457,
+      "step": 3811
+    },
+    {
+      "epoch": 0.8588728983017433,
+      "grad_norm": 0.47847573245900543,
+      "learning_rate": 2.6709631513851834e-07,
+      "loss": 0.1422,
+      "step": 3812
+    },
+    {
+      "epoch": 0.8590982059875517,
+      "grad_norm": 0.4729488813409411,
+      "learning_rate": 2.6625912222267844e-07,
+      "loss": 0.1464,
+      "step": 3813
+    },
+    {
+      "epoch": 0.8593235136733601,
+      "grad_norm": 0.4742682819045076,
+      "learning_rate": 2.654231696198878e-07,
+      "loss": 0.1394,
+      "step": 3814
+    },
+    {
+      "epoch": 0.8595488213591687,
+      "grad_norm": 0.4496738891729263,
+      "learning_rate": 2.645884577943192e-07,
+      "loss": 0.1308,
+      "step": 3815
+    },
+    {
+      "epoch": 0.8597741290449771,
+      "grad_norm": 0.4857654847446031,
+      "learning_rate": 2.6375498720945717e-07,
+      "loss": 0.1457,
+      "step": 3816
+    },
+    {
+      "epoch": 0.8599994367307855,
+      "grad_norm": 0.4835247667742374,
+      "learning_rate": 2.629227583280972e-07,
+      "loss": 0.1408,
+      "step": 3817
+    },
+    {
+      "epoch": 0.8602247444165939,
+      "grad_norm": 0.512090740233511,
+      "learning_rate": 2.620917716123444e-07,
+      "loss": 0.1638,
+      "step": 3818
+    },
+    {
+      "epoch": 0.8604500521024023,
+      "grad_norm": 0.5050589725161969,
+      "learning_rate": 2.6126202752361554e-07,
+      "loss": 0.1475,
+      "step": 3819
+    },
+    {
+      "epoch": 0.8606753597882107,
+      "grad_norm": 0.4797067731221806,
+      "learning_rate": 2.604335265226354e-07,
+      "loss": 0.138,
+      "step": 3820
+    },
+    {
+      "epoch": 0.8609006674740192,
+      "grad_norm": 0.47641645566421503,
+      "learning_rate": 2.5960626906944066e-07,
+      "loss": 0.1429,
+      "step": 3821
+    },
+    {
+      "epoch": 0.8611259751598277,
+      "grad_norm": 0.4949453699572789,
+      "learning_rate": 2.587802556233765e-07,
+      "loss": 0.1489,
+      "step": 3822
+    },
+    {
+      "epoch": 0.8613512828456361,
+      "grad_norm": 0.46752972020335043,
+      "learning_rate": 2.5795548664309695e-07,
+      "loss": 0.1362,
+      "step": 3823
+    },
+    {
+      "epoch": 0.8615765905314445,
+      "grad_norm": 0.4617599130698046,
+      "learning_rate": 2.571319625865662e-07,
+      "loss": 0.1379,
+      "step": 3824
+    },
+    {
+      "epoch": 0.861801898217253,
+      "grad_norm": 0.46615766557140464,
+      "learning_rate": 2.5630968391105515e-07,
+      "loss": 0.1458,
+      "step": 3825
+    },
+    {
+      "epoch": 0.8620272059030614,
+      "grad_norm": 0.4713775308849494,
+      "learning_rate": 2.5548865107314606e-07,
+      "loss": 0.1344,
+      "step": 3826
+    },
+    {
+      "epoch": 0.8622525135888698,
+      "grad_norm": 0.4718354741536646,
+      "learning_rate": 2.546688645287268e-07,
+      "loss": 0.1346,
+      "step": 3827
+    },
+    {
+      "epoch": 0.8624778212746782,
+      "grad_norm": 0.4762759961692442,
+      "learning_rate": 2.5385032473299433e-07,
+      "loss": 0.1403,
+      "step": 3828
+    },
+    {
+      "epoch": 0.8627031289604866,
+      "grad_norm": 0.45935519635317734,
+      "learning_rate": 2.5303303214045423e-07,
+      "loss": 0.1381,
+      "step": 3829
+    },
+    {
+      "epoch": 0.8629284366462951,
+      "grad_norm": 0.4890217885035087,
+      "learning_rate": 2.522169872049174e-07,
+      "loss": 0.1565,
+      "step": 3830
+    },
+    {
+      "epoch": 0.8631537443321036,
+      "grad_norm": 0.4647472157245905,
+      "learning_rate": 2.5140219037950416e-07,
+      "loss": 0.1424,
+      "step": 3831
+    },
+    {
+      "epoch": 0.863379052017912,
+      "grad_norm": 0.5084070140741922,
+      "learning_rate": 2.5058864211664064e-07,
+      "loss": 0.1532,
+      "step": 3832
+    },
+    {
+      "epoch": 0.8636043597037204,
+      "grad_norm": 0.5116273779430994,
+      "learning_rate": 2.4977634286805887e-07,
+      "loss": 0.1622,
+      "step": 3833
+    },
+    {
+      "epoch": 0.8638296673895288,
+      "grad_norm": 0.47251055408026604,
+      "learning_rate": 2.4896529308479966e-07,
+      "loss": 0.1394,
+      "step": 3834
+    },
+    {
+      "epoch": 0.8640549750753372,
+      "grad_norm": 0.45085871056941923,
+      "learning_rate": 2.4815549321720755e-07,
+      "loss": 0.1309,
+      "step": 3835
+    },
+    {
+      "epoch": 0.8642802827611457,
+      "grad_norm": 0.45540371760054993,
+      "learning_rate": 2.4734694371493507e-07,
+      "loss": 0.1301,
+      "step": 3836
+    },
+    {
+      "epoch": 0.8645055904469541,
+      "grad_norm": 0.4601648252123752,
+      "learning_rate": 2.4653964502693974e-07,
+      "loss": 0.137,
+      "step": 3837
+    },
+    {
+      "epoch": 0.8647308981327626,
+      "grad_norm": 0.48395343387323725,
+      "learning_rate": 2.4573359760148354e-07,
+      "loss": 0.1418,
+      "step": 3838
+    },
+    {
+      "epoch": 0.864956205818571,
+      "grad_norm": 0.466410678978402,
+      "learning_rate": 2.449288018861354e-07,
+      "loss": 0.1443,
+      "step": 3839
+    },
+    {
+      "epoch": 0.8651815135043794,
+      "grad_norm": 0.4594105233412585,
+      "learning_rate": 2.441252583277678e-07,
+      "loss": 0.1329,
+      "step": 3840
+    },
+    {
+      "epoch": 0.8654068211901879,
+      "grad_norm": 0.4645514686858519,
+      "learning_rate": 2.433229673725593e-07,
+      "loss": 0.1309,
+      "step": 3841
+    },
+    {
+      "epoch": 0.8656321288759963,
+      "grad_norm": 0.4862741309606466,
+      "learning_rate": 2.425219294659908e-07,
+      "loss": 0.1497,
+      "step": 3842
+    },
+    {
+      "epoch": 0.8658574365618047,
+      "grad_norm": 0.5030380126363495,
+      "learning_rate": 2.4172214505285006e-07,
+      "loss": 0.1501,
+      "step": 3843
+    },
+    {
+      "epoch": 0.8660827442476131,
+      "grad_norm": 0.47748877151047414,
+      "learning_rate": 2.409236145772276e-07,
+      "loss": 0.137,
+      "step": 3844
+    },
+    {
+      "epoch": 0.8663080519334215,
+      "grad_norm": 0.46467065040499544,
+      "learning_rate": 2.401263384825164e-07,
+      "loss": 0.1382,
+      "step": 3845
+    },
+    {
+      "epoch": 0.8665333596192301,
+      "grad_norm": 0.464481683671317,
+      "learning_rate": 2.393303172114159e-07,
+      "loss": 0.1337,
+      "step": 3846
+    },
+    {
+      "epoch": 0.8667586673050385,
+      "grad_norm": 0.46035212211555143,
+      "learning_rate": 2.3853555120592506e-07,
+      "loss": 0.13,
+      "step": 3847
+    },
+    {
+      "epoch": 0.8669839749908469,
+      "grad_norm": 0.4637142538065829,
+      "learning_rate": 2.377420409073497e-07,
+      "loss": 0.1331,
+      "step": 3848
+    },
+    {
+      "epoch": 0.8672092826766553,
+      "grad_norm": 0.4632361803177915,
+      "learning_rate": 2.3694978675629476e-07,
+      "loss": 0.1452,
+      "step": 3849
+    },
+    {
+      "epoch": 0.8674345903624637,
+      "grad_norm": 0.4986572200737204,
+      "learning_rate": 2.3615878919267116e-07,
+      "loss": 0.1577,
+      "step": 3850
+    },
+    {
+      "epoch": 0.8676598980482721,
+      "grad_norm": 0.4914501429064226,
+      "learning_rate": 2.3536904865568949e-07,
+      "loss": 0.1461,
+      "step": 3851
+    },
+    {
+      "epoch": 0.8678852057340806,
+      "grad_norm": 0.46554246389297854,
+      "learning_rate": 2.345805655838626e-07,
+      "loss": 0.1444,
+      "step": 3852
+    },
+    {
+      "epoch": 0.868110513419889,
+      "grad_norm": 0.49645474855635946,
+      "learning_rate": 2.337933404150064e-07,
+      "loss": 0.1579,
+      "step": 3853
+    },
+    {
+      "epoch": 0.8683358211056975,
+      "grad_norm": 0.47708434402385985,
+      "learning_rate": 2.3300737358623843e-07,
+      "loss": 0.1497,
+      "step": 3854
+    },
+    {
+      "epoch": 0.8685611287915059,
+      "grad_norm": 0.46968566557594305,
+      "learning_rate": 2.3222266553397542e-07,
+      "loss": 0.1492,
+      "step": 3855
+    },
+    {
+      "epoch": 0.8687864364773143,
+      "grad_norm": 0.4878675904608552,
+      "learning_rate": 2.314392166939375e-07,
+      "loss": 0.1471,
+      "step": 3856
+    },
+    {
+      "epoch": 0.8690117441631228,
+      "grad_norm": 0.45020022653810526,
+      "learning_rate": 2.3065702750114383e-07,
+      "loss": 0.1292,
+      "step": 3857
+    },
+    {
+      "epoch": 0.8692370518489312,
+      "grad_norm": 0.4843459055982729,
+      "learning_rate": 2.2987609838991536e-07,
+      "loss": 0.1387,
+      "step": 3858
+    },
+    {
+      "epoch": 0.8694623595347396,
+      "grad_norm": 0.47005449497641894,
+      "learning_rate": 2.2909642979387331e-07,
+      "loss": 0.1445,
+      "step": 3859
+    },
+    {
+      "epoch": 0.869687667220548,
+      "grad_norm": 0.49344268598068725,
+      "learning_rate": 2.2831802214593774e-07,
+      "loss": 0.1513,
+      "step": 3860
+    },
+    {
+      "epoch": 0.8699129749063564,
+      "grad_norm": 0.48046850809610137,
+      "learning_rate": 2.2754087587833014e-07,
+      "loss": 0.136,
+      "step": 3861
+    },
+    {
+      "epoch": 0.870138282592165,
+      "grad_norm": 0.5006165000856634,
+      "learning_rate": 2.2676499142257002e-07,
+      "loss": 0.1596,
+      "step": 3862
+    },
+    {
+      "epoch": 0.8703635902779734,
+      "grad_norm": 0.47972327496783845,
+      "learning_rate": 2.2599036920947836e-07,
+      "loss": 0.1416,
+      "step": 3863
+    },
+    {
+      "epoch": 0.8705888979637818,
+      "grad_norm": 0.47816719003504604,
+      "learning_rate": 2.2521700966917276e-07,
+      "loss": 0.1361,
+      "step": 3864
+    },
+    {
+      "epoch": 0.8708142056495902,
+      "grad_norm": 0.4664317470965975,
+      "learning_rate": 2.2444491323107138e-07,
+      "loss": 0.1368,
+      "step": 3865
+    },
+    {
+      "epoch": 0.8710395133353986,
+      "grad_norm": 0.46907919230886,
+      "learning_rate": 2.23674080323891e-07,
+      "loss": 0.143,
+      "step": 3866
+    },
+    {
+      "epoch": 0.8712648210212071,
+      "grad_norm": 0.48697418501853557,
+      "learning_rate": 2.229045113756456e-07,
+      "loss": 0.1542,
+      "step": 3867
+    },
+    {
+      "epoch": 0.8714901287070155,
+      "grad_norm": 0.5031905886611365,
+      "learning_rate": 2.221362068136493e-07,
+      "loss": 0.1484,
+      "step": 3868
+    },
+    {
+      "epoch": 0.8717154363928239,
+      "grad_norm": 0.5014679159314208,
+      "learning_rate": 2.2136916706451212e-07,
+      "loss": 0.1432,
+      "step": 3869
+    },
+    {
+      "epoch": 0.8719407440786324,
+      "grad_norm": 0.48675398397776903,
+      "learning_rate": 2.2060339255414232e-07,
+      "loss": 0.1448,
+      "step": 3870
+    },
+    {
+      "epoch": 0.8721660517644408,
+      "grad_norm": 0.4719989893940155,
+      "learning_rate": 2.198388837077467e-07,
+      "loss": 0.1399,
+      "step": 3871
+    },
+    {
+      "epoch": 0.8723913594502493,
+      "grad_norm": 0.4698672166182498,
+      "learning_rate": 2.190756409498282e-07,
+      "loss": 0.1432,
+      "step": 3872
+    },
+    {
+      "epoch": 0.8726166671360577,
+      "grad_norm": 0.5003866524002453,
+      "learning_rate": 2.1831366470418725e-07,
+      "loss": 0.1597,
+      "step": 3873
+    },
+    {
+      "epoch": 0.8728419748218661,
+      "grad_norm": 0.4762260917827092,
+      "learning_rate": 2.175529553939204e-07,
+      "loss": 0.1407,
+      "step": 3874
+    },
+    {
+      "epoch": 0.8730672825076745,
+      "grad_norm": 0.46479989802739047,
+      "learning_rate": 2.1679351344142146e-07,
+      "loss": 0.1342,
+      "step": 3875
+    },
+    {
+      "epoch": 0.8732925901934829,
+      "grad_norm": 0.4739375337047061,
+      "learning_rate": 2.1603533926838088e-07,
+      "loss": 0.1481,
+      "step": 3876
+    },
+    {
+      "epoch": 0.8735178978792915,
+      "grad_norm": 0.46262128348097714,
+      "learning_rate": 2.1527843329578328e-07,
+      "loss": 0.1402,
+      "step": 3877
+    },
+    {
+      "epoch": 0.8737432055650999,
+      "grad_norm": 0.5027757553131618,
+      "learning_rate": 2.1452279594391167e-07,
+      "loss": 0.1517,
+      "step": 3878
+    },
+    {
+      "epoch": 0.8739685132509083,
+      "grad_norm": 0.4552996612527901,
+      "learning_rate": 2.1376842763234178e-07,
+      "loss": 0.1328,
+      "step": 3879
+    },
+    {
+      "epoch": 0.8741938209367167,
+      "grad_norm": 0.4554471045600218,
+      "learning_rate": 2.1301532877994747e-07,
+      "loss": 0.1294,
+      "step": 3880
+    },
+    {
+      "epoch": 0.8744191286225251,
+      "grad_norm": 0.49367764827210675,
+      "learning_rate": 2.1226349980489614e-07,
+      "loss": 0.1566,
+      "step": 3881
+    },
+    {
+      "epoch": 0.8746444363083336,
+      "grad_norm": 0.4661026591919418,
+      "learning_rate": 2.1151294112464997e-07,
+      "loss": 0.1416,
+      "step": 3882
+    },
+    {
+      "epoch": 0.874869743994142,
+      "grad_norm": 0.4920345042810052,
+      "learning_rate": 2.1076365315596704e-07,
+      "loss": 0.153,
+      "step": 3883
+    },
+    {
+      "epoch": 0.8750950516799504,
+      "grad_norm": 0.4789670218119234,
+      "learning_rate": 2.1001563631489807e-07,
+      "loss": 0.1434,
+      "step": 3884
+    },
+    {
+      "epoch": 0.8753203593657589,
+      "grad_norm": 0.4716950050495697,
+      "learning_rate": 2.0926889101679004e-07,
+      "loss": 0.1387,
+      "step": 3885
+    },
+    {
+      "epoch": 0.8755456670515673,
+      "grad_norm": 0.47883106851183405,
+      "learning_rate": 2.0852341767628182e-07,
+      "loss": 0.1496,
+      "step": 3886
+    },
+    {
+      "epoch": 0.8757709747373758,
+      "grad_norm": 0.4959167987665322,
+      "learning_rate": 2.07779216707307e-07,
+      "loss": 0.151,
+      "step": 3887
+    },
+    {
+      "epoch": 0.8759962824231842,
+      "grad_norm": 0.47334563856241746,
+      "learning_rate": 2.0703628852309336e-07,
+      "loss": 0.1407,
+      "step": 3888
+    },
+    {
+      "epoch": 0.8762215901089926,
+      "grad_norm": 0.4627588097838559,
+      "learning_rate": 2.0629463353616013e-07,
+      "loss": 0.1437,
+      "step": 3889
+    },
+    {
+      "epoch": 0.876446897794801,
+      "grad_norm": 0.48615089687091206,
+      "learning_rate": 2.0555425215832176e-07,
+      "loss": 0.1352,
+      "step": 3890
+    },
+    {
+      "epoch": 0.8766722054806094,
+      "grad_norm": 0.4911026246700992,
+      "learning_rate": 2.048151448006841e-07,
+      "loss": 0.1551,
+      "step": 3891
+    },
+    {
+      "epoch": 0.8768975131664178,
+      "grad_norm": 0.46591980702852437,
+      "learning_rate": 2.0407731187364556e-07,
+      "loss": 0.1416,
+      "step": 3892
+    },
+    {
+      "epoch": 0.8771228208522264,
+      "grad_norm": 0.49095474891414137,
+      "learning_rate": 2.0334075378689781e-07,
+      "loss": 0.1534,
+      "step": 3893
+    },
+    {
+      "epoch": 0.8773481285380348,
+      "grad_norm": 0.4667276709190779,
+      "learning_rate": 2.026054709494235e-07,
+      "loss": 0.1446,
+      "step": 3894
+    },
+    {
+      "epoch": 0.8775734362238432,
+      "grad_norm": 0.47067318796567426,
+      "learning_rate": 2.0187146376949852e-07,
+      "loss": 0.1362,
+      "step": 3895
+    },
+    {
+      "epoch": 0.8777987439096516,
+      "grad_norm": 0.5010169634897643,
+      "learning_rate": 2.0113873265468875e-07,
+      "loss": 0.1497,
+      "step": 3896
+    },
+    {
+      "epoch": 0.87802405159546,
+      "grad_norm": 0.47896885342438095,
+      "learning_rate": 2.0040727801185323e-07,
+      "loss": 0.1404,
+      "step": 3897
+    },
+    {
+      "epoch": 0.8782493592812685,
+      "grad_norm": 0.4435881047514352,
+      "learning_rate": 1.996771002471415e-07,
+      "loss": 0.13,
+      "step": 3898
+    },
+    {
+      "epoch": 0.8784746669670769,
+      "grad_norm": 0.46596846807949904,
+      "learning_rate": 1.9894819976599338e-07,
+      "loss": 0.1377,
+      "step": 3899
+    },
+    {
+      "epoch": 0.8786999746528853,
+      "grad_norm": 0.4644431447384091,
+      "learning_rate": 1.9822057697314102e-07,
+      "loss": 0.1326,
+      "step": 3900
+    },
+    {
+      "epoch": 0.8789252823386938,
+      "grad_norm": 0.4672389855441275,
+      "learning_rate": 1.9749423227260533e-07,
+      "loss": 0.1299,
+      "step": 3901
+    },
+    {
+      "epoch": 0.8791505900245022,
+      "grad_norm": 0.4686815604792627,
+      "learning_rate": 1.9676916606769874e-07,
+      "loss": 0.1469,
+      "step": 3902
+    },
+    {
+      "epoch": 0.8793758977103107,
+      "grad_norm": 0.4659336099421951,
+      "learning_rate": 1.9604537876102448e-07,
+      "loss": 0.1398,
+      "step": 3903
+    },
+    {
+      "epoch": 0.8796012053961191,
+      "grad_norm": 0.443210123551164,
+      "learning_rate": 1.9532287075447325e-07,
+      "loss": 0.1341,
+      "step": 3904
+    },
+    {
+      "epoch": 0.8798265130819275,
+      "grad_norm": 0.47878160196488945,
+      "learning_rate": 1.9460164244922698e-07,
+      "loss": 0.1429,
+      "step": 3905
+    },
+    {
+      "epoch": 0.8800518207677359,
+      "grad_norm": 0.48539640117019267,
+      "learning_rate": 1.9388169424575802e-07,
+      "loss": 0.14,
+      "step": 3906
+    },
+    {
+      "epoch": 0.8802771284535443,
+      "grad_norm": 0.46434475314245555,
+      "learning_rate": 1.9316302654382528e-07,
+      "loss": 0.1337,
+      "step": 3907
+    },
+    {
+      "epoch": 0.8805024361393528,
+      "grad_norm": 0.4851787509947083,
+      "learning_rate": 1.9244563974247953e-07,
+      "loss": 0.1489,
+      "step": 3908
+    },
+    {
+      "epoch": 0.8807277438251613,
+      "grad_norm": 0.4911309040191241,
+      "learning_rate": 1.917295342400577e-07,
+      "loss": 0.1497,
+      "step": 3909
+    },
+    {
+      "epoch": 0.8809530515109697,
+      "grad_norm": 0.45607420361153655,
+      "learning_rate": 1.910147104341875e-07,
+      "loss": 0.1301,
+      "step": 3910
+    },
+    {
+      "epoch": 0.8811783591967781,
+      "grad_norm": 0.47282797295865775,
+      "learning_rate": 1.9030116872178317e-07,
+      "loss": 0.1537,
+      "step": 3911
+    },
+    {
+      "epoch": 0.8814036668825865,
+      "grad_norm": 0.5049449835303502,
+      "learning_rate": 1.8958890949904802e-07,
+      "loss": 0.1581,
+      "step": 3912
+    },
+    {
+      "epoch": 0.881628974568395,
+      "grad_norm": 0.47520130714031167,
+      "learning_rate": 1.8887793316147373e-07,
+      "loss": 0.1397,
+      "step": 3913
+    },
+    {
+      "epoch": 0.8818542822542034,
+      "grad_norm": 0.48093307567540367,
+      "learning_rate": 1.881682401038379e-07,
+      "loss": 0.1369,
+      "step": 3914
+    },
+    {
+      "epoch": 0.8820795899400118,
+      "grad_norm": 0.4374287939212524,
+      "learning_rate": 1.8745983072020774e-07,
+      "loss": 0.1315,
+      "step": 3915
+    },
+    {
+      "epoch": 0.8823048976258202,
+      "grad_norm": 0.501261525882023,
+      "learning_rate": 1.8675270540393532e-07,
+      "loss": 0.1461,
+      "step": 3916
+    },
+    {
+      "epoch": 0.8825302053116287,
+      "grad_norm": 0.4878953885886656,
+      "learning_rate": 1.8604686454766208e-07,
+      "loss": 0.152,
+      "step": 3917
+    },
+    {
+      "epoch": 0.8827555129974372,
+      "grad_norm": 0.47618468982397694,
+      "learning_rate": 1.8534230854331454e-07,
+      "loss": 0.1473,
+      "step": 3918
+    },
+    {
+      "epoch": 0.8829808206832456,
+      "grad_norm": 0.48255019165112745,
+      "learning_rate": 1.8463903778210612e-07,
+      "loss": 0.1525,
+      "step": 3919
+    },
+    {
+      "epoch": 0.883206128369054,
+      "grad_norm": 0.4709200797557006,
+      "learning_rate": 1.8393705265453838e-07,
+      "loss": 0.1445,
+      "step": 3920
+    },
+    {
+      "epoch": 0.8834314360548624,
+      "grad_norm": 0.4726612900241033,
+      "learning_rate": 1.832363535503956e-07,
+      "loss": 0.1465,
+      "step": 3921
+    },
+    {
+      "epoch": 0.8836567437406708,
+      "grad_norm": 0.4648816875604027,
+      "learning_rate": 1.8253694085875047e-07,
+      "loss": 0.1289,
+      "step": 3922
+    },
+    {
+      "epoch": 0.8838820514264792,
+      "grad_norm": 0.49096699990870324,
+      "learning_rate": 1.8183881496796146e-07,
+      "loss": 0.1502,
+      "step": 3923
+    },
+    {
+      "epoch": 0.8841073591122878,
+      "grad_norm": 0.5020026232845864,
+      "learning_rate": 1.8114197626567105e-07,
+      "loss": 0.1589,
+      "step": 3924
+    },
+    {
+      "epoch": 0.8843326667980962,
+      "grad_norm": 0.5030844197467426,
+      "learning_rate": 1.8044642513880827e-07,
+      "loss": 0.1557,
+      "step": 3925
+    },
+    {
+      "epoch": 0.8845579744839046,
+      "grad_norm": 0.48572411012187333,
+      "learning_rate": 1.7975216197358648e-07,
+      "loss": 0.1422,
+      "step": 3926
+    },
+    {
+      "epoch": 0.884783282169713,
+      "grad_norm": 0.47185579135933253,
+      "learning_rate": 1.790591871555039e-07,
+      "loss": 0.1424,
+      "step": 3927
+    },
+    {
+      "epoch": 0.8850085898555214,
+      "grad_norm": 0.45572123796883646,
+      "learning_rate": 1.7836750106934475e-07,
+      "loss": 0.1333,
+      "step": 3928
+    },
+    {
+      "epoch": 0.8852338975413299,
+      "grad_norm": 0.4777374273399248,
+      "learning_rate": 1.776771040991751e-07,
+      "loss": 0.1381,
+      "step": 3929
+    },
+    {
+      "epoch": 0.8854592052271383,
+      "grad_norm": 0.49018085298222697,
+      "learning_rate": 1.7698799662834776e-07,
+      "loss": 0.1467,
+      "step": 3930
+    },
+    {
+      "epoch": 0.8856845129129467,
+      "grad_norm": 0.47340916213212886,
+      "learning_rate": 1.7630017903949775e-07,
+      "loss": 0.1451,
+      "step": 3931
+    },
+    {
+      "epoch": 0.8859098205987552,
+      "grad_norm": 0.49233987363836046,
+      "learning_rate": 1.7561365171454488e-07,
+      "loss": 0.1507,
+      "step": 3932
+    },
+    {
+      "epoch": 0.8861351282845636,
+      "grad_norm": 0.44864984855065165,
+      "learning_rate": 1.7492841503469165e-07,
+      "loss": 0.136,
+      "step": 3933
+    },
+    {
+      "epoch": 0.8863604359703721,
+      "grad_norm": 0.46498551221740236,
+      "learning_rate": 1.7424446938042517e-07,
+      "loss": 0.1435,
+      "step": 3934
+    },
+    {
+      "epoch": 0.8865857436561805,
+      "grad_norm": 0.494552407996217,
+      "learning_rate": 1.7356181513151464e-07,
+      "loss": 0.1347,
+      "step": 3935
+    },
+    {
+      "epoch": 0.8868110513419889,
+      "grad_norm": 0.46218453673553506,
+      "learning_rate": 1.7288045266701247e-07,
+      "loss": 0.1271,
+      "step": 3936
+    },
+    {
+      "epoch": 0.8870363590277973,
+      "grad_norm": 0.4620631245597858,
+      "learning_rate": 1.7220038236525406e-07,
+      "loss": 0.1411,
+      "step": 3937
+    },
+    {
+      "epoch": 0.8872616667136057,
+      "grad_norm": 0.5238585649293778,
+      "learning_rate": 1.7152160460385703e-07,
+      "loss": 0.1372,
+      "step": 3938
+    },
+    {
+      "epoch": 0.8874869743994142,
+      "grad_norm": 0.4980800570271065,
+      "learning_rate": 1.7084411975972076e-07,
+      "loss": 0.1514,
+      "step": 3939
+    },
+    {
+      "epoch": 0.8877122820852227,
+      "grad_norm": 0.4779022749671553,
+      "learning_rate": 1.701679282090285e-07,
+      "loss": 0.1486,
+      "step": 3940
+    },
+    {
+      "epoch": 0.8879375897710311,
+      "grad_norm": 0.4723075786530799,
+      "learning_rate": 1.6949303032724297e-07,
+      "loss": 0.1373,
+      "step": 3941
+    },
+    {
+      "epoch": 0.8881628974568395,
+      "grad_norm": 0.4787812903509886,
+      "learning_rate": 1.6881942648911077e-07,
+      "loss": 0.1341,
+      "step": 3942
+    },
+    {
+      "epoch": 0.8883882051426479,
+      "grad_norm": 0.49509450884332845,
+      "learning_rate": 1.6814711706865827e-07,
+      "loss": 0.1512,
+      "step": 3943
+    },
+    {
+      "epoch": 0.8886135128284564,
+      "grad_norm": 0.48155729793324203,
+      "learning_rate": 1.6747610243919437e-07,
+      "loss": 0.1398,
+      "step": 3944
+    },
+    {
+      "epoch": 0.8888388205142648,
+      "grad_norm": 0.4859521054322674,
+      "learning_rate": 1.6680638297330854e-07,
+      "loss": 0.1456,
+      "step": 3945
+    },
+    {
+      "epoch": 0.8890641282000732,
+      "grad_norm": 0.4922788181816615,
+      "learning_rate": 1.661379590428705e-07,
+      "loss": 0.1546,
+      "step": 3946
+    },
+    {
+      "epoch": 0.8892894358858816,
+      "grad_norm": 0.4720865220150437,
+      "learning_rate": 1.6547083101903173e-07,
+      "loss": 0.1407,
+      "step": 3947
+    },
+    {
+      "epoch": 0.8895147435716901,
+      "grad_norm": 0.5004812140336558,
+      "learning_rate": 1.6480499927222283e-07,
+      "loss": 0.1535,
+      "step": 3948
+    },
+    {
+      "epoch": 0.8897400512574986,
+      "grad_norm": 0.44759678855803414,
+      "learning_rate": 1.641404641721561e-07,
+      "loss": 0.1288,
+      "step": 3949
+    },
+    {
+      "epoch": 0.889965358943307,
+      "grad_norm": 0.48781988498782375,
+      "learning_rate": 1.6347722608782284e-07,
+      "loss": 0.1499,
+      "step": 3950
+    },
+    {
+      "epoch": 0.8901906666291154,
+      "grad_norm": 0.4383912218136226,
+      "learning_rate": 1.6281528538749425e-07,
+      "loss": 0.1324,
+      "step": 3951
+    },
+    {
+      "epoch": 0.8904159743149238,
+      "grad_norm": 0.4598988414847289,
+      "learning_rate": 1.6215464243872186e-07,
+      "loss": 0.1333,
+      "step": 3952
+    },
+    {
+      "epoch": 0.8906412820007322,
+      "grad_norm": 0.5021102204148211,
+      "learning_rate": 1.6149529760833504e-07,
+      "loss": 0.148,
+      "step": 3953
+    },
+    {
+      "epoch": 0.8908665896865406,
+      "grad_norm": 0.43661485835748093,
+      "learning_rate": 1.608372512624448e-07,
+      "loss": 0.1243,
+      "step": 3954
+    },
+    {
+      "epoch": 0.8910918973723491,
+      "grad_norm": 0.45412666194626644,
+      "learning_rate": 1.6018050376643863e-07,
+      "loss": 0.1348,
+      "step": 3955
+    },
+    {
+      "epoch": 0.8913172050581576,
+      "grad_norm": 0.4703275266482927,
+      "learning_rate": 1.595250554849842e-07,
+      "loss": 0.1433,
+      "step": 3956
+    },
+    {
+      "epoch": 0.891542512743966,
+      "grad_norm": 0.47470587203113024,
+      "learning_rate": 1.5887090678202793e-07,
+      "loss": 0.1404,
+      "step": 3957
+    },
+    {
+      "epoch": 0.8917678204297744,
+      "grad_norm": 0.4754627455351376,
+      "learning_rate": 1.5821805802079343e-07,
+      "loss": 0.141,
+      "step": 3958
+    },
+    {
+      "epoch": 0.8919931281155828,
+      "grad_norm": 0.46616519097330594,
+      "learning_rate": 1.5756650956378377e-07,
+      "loss": 0.1485,
+      "step": 3959
+    },
+    {
+      "epoch": 0.8922184358013913,
+      "grad_norm": 0.461205924770751,
+      "learning_rate": 1.5691626177277986e-07,
+      "loss": 0.1416,
+      "step": 3960
+    },
+    {
+      "epoch": 0.8924437434871997,
+      "grad_norm": 0.47273107411957194,
+      "learning_rate": 1.5626731500883951e-07,
+      "loss": 0.1411,
+      "step": 3961
+    },
+    {
+      "epoch": 0.8926690511730081,
+      "grad_norm": 0.48751682609530655,
+      "learning_rate": 1.5561966963229925e-07,
+      "loss": 0.1438,
+      "step": 3962
+    },
+    {
+      "epoch": 0.8928943588588165,
+      "grad_norm": 0.46850837700229225,
+      "learning_rate": 1.5497332600277137e-07,
+      "loss": 0.1418,
+      "step": 3963
+    },
+    {
+      "epoch": 0.893119666544625,
+      "grad_norm": 0.45741592189608293,
+      "learning_rate": 1.5432828447914743e-07,
+      "loss": 0.143,
+      "step": 3964
+    },
+    {
+      "epoch": 0.8933449742304335,
+      "grad_norm": 0.4616383354901837,
+      "learning_rate": 1.5368454541959453e-07,
+      "loss": 0.1324,
+      "step": 3965
+    },
+    {
+      "epoch": 0.8935702819162419,
+      "grad_norm": 0.5020609237394296,
+      "learning_rate": 1.5304210918155677e-07,
+      "loss": 0.1537,
+      "step": 3966
+    },
+    {
+      "epoch": 0.8937955896020503,
+      "grad_norm": 0.4568905419965394,
+      "learning_rate": 1.524009761217557e-07,
+      "loss": 0.1378,
+      "step": 3967
+    },
+    {
+      "epoch": 0.8940208972878587,
+      "grad_norm": 0.46651422878987997,
+      "learning_rate": 1.5176114659618796e-07,
+      "loss": 0.1406,
+      "step": 3968
+    },
+    {
+      "epoch": 0.8942462049736671,
+      "grad_norm": 0.4667624257365965,
+      "learning_rate": 1.5112262096012743e-07,
+      "loss": 0.1356,
+      "step": 3969
+    },
+    {
+      "epoch": 0.8944715126594756,
+      "grad_norm": 0.46212811202290244,
+      "learning_rate": 1.5048539956812324e-07,
+      "loss": 0.141,
+      "step": 3970
+    },
+    {
+      "epoch": 0.8946968203452841,
+      "grad_norm": 0.4854084542243427,
+      "learning_rate": 1.4984948277400074e-07,
+      "loss": 0.1496,
+      "step": 3971
+    },
+    {
+      "epoch": 0.8949221280310925,
+      "grad_norm": 0.46644509385921373,
+      "learning_rate": 1.4921487093086134e-07,
+      "loss": 0.1478,
+      "step": 3972
+    },
+    {
+      "epoch": 0.8951474357169009,
+      "grad_norm": 0.4593367044330634,
+      "learning_rate": 1.4858156439108097e-07,
+      "loss": 0.1269,
+      "step": 3973
+    },
+    {
+      "epoch": 0.8953727434027093,
+      "grad_norm": 0.48517550690122074,
+      "learning_rate": 1.4794956350631106e-07,
+      "loss": 0.15,
+      "step": 3974
+    },
+    {
+      "epoch": 0.8955980510885178,
+      "grad_norm": 0.4808142374337991,
+      "learning_rate": 1.473188686274782e-07,
+      "loss": 0.1421,
+      "step": 3975
+    },
+    {
+      "epoch": 0.8958233587743262,
+      "grad_norm": 0.49699266447125817,
+      "learning_rate": 1.4668948010478358e-07,
+      "loss": 0.1486,
+      "step": 3976
+    },
+    {
+      "epoch": 0.8960486664601346,
+      "grad_norm": 0.4718085033512432,
+      "learning_rate": 1.4606139828770378e-07,
+      "loss": 0.1481,
+      "step": 3977
+    },
+    {
+      "epoch": 0.896273974145943,
+      "grad_norm": 0.480493817674567,
+      "learning_rate": 1.4543462352498844e-07,
+      "loss": 0.1378,
+      "step": 3978
+    },
+    {
+      "epoch": 0.8964992818317515,
+      "grad_norm": 0.44881325605798783,
+      "learning_rate": 1.448091561646628e-07,
+      "loss": 0.1418,
+      "step": 3979
+    },
+    {
+      "epoch": 0.89672458951756,
+      "grad_norm": 0.45855300593285897,
+      "learning_rate": 1.4418499655402512e-07,
+      "loss": 0.1333,
+      "step": 3980
+    },
+    {
+      "epoch": 0.8969498972033684,
+      "grad_norm": 0.4667832417601448,
+      "learning_rate": 1.435621450396485e-07,
+      "loss": 0.142,
+      "step": 3981
+    },
+    {
+      "epoch": 0.8971752048891768,
+      "grad_norm": 0.4861168968624223,
+      "learning_rate": 1.4294060196737874e-07,
+      "loss": 0.1379,
+      "step": 3982
+    },
+    {
+      "epoch": 0.8974005125749852,
+      "grad_norm": 0.5123715647299666,
+      "learning_rate": 1.4232036768233565e-07,
+      "loss": 0.1524,
+      "step": 3983
+    },
+    {
+      "epoch": 0.8976258202607936,
+      "grad_norm": 0.46111866378689154,
+      "learning_rate": 1.417014425289126e-07,
+      "loss": 0.1382,
+      "step": 3984
+    },
+    {
+      "epoch": 0.897851127946602,
+      "grad_norm": 0.46962766756276614,
+      "learning_rate": 1.4108382685077498e-07,
+      "loss": 0.138,
+      "step": 3985
+    },
+    {
+      "epoch": 0.8980764356324105,
+      "grad_norm": 0.4868405824344992,
+      "learning_rate": 1.4046752099086236e-07,
+      "loss": 0.1451,
+      "step": 3986
+    },
+    {
+      "epoch": 0.898301743318219,
+      "grad_norm": 0.4836538182472601,
+      "learning_rate": 1.398525252913857e-07,
+      "loss": 0.1465,
+      "step": 3987
+    },
+    {
+      "epoch": 0.8985270510040274,
+      "grad_norm": 0.4552388575669367,
+      "learning_rate": 1.3923884009382994e-07,
+      "loss": 0.1383,
+      "step": 3988
+    },
+    {
+      "epoch": 0.8987523586898358,
+      "grad_norm": 0.46674175236849347,
+      "learning_rate": 1.3862646573895134e-07,
+      "loss": 0.1465,
+      "step": 3989
+    },
+    {
+      "epoch": 0.8989776663756442,
+      "grad_norm": 0.46416844608939184,
+      "learning_rate": 1.380154025667782e-07,
+      "loss": 0.139,
+      "step": 3990
+    },
+    {
+      "epoch": 0.8992029740614527,
+      "grad_norm": 0.48341279291397105,
+      "learning_rate": 1.374056509166119e-07,
+      "loss": 0.146,
+      "step": 3991
+    },
+    {
+      "epoch": 0.8994282817472611,
+      "grad_norm": 0.44154753930596685,
+      "learning_rate": 1.367972111270241e-07,
+      "loss": 0.1264,
+      "step": 3992
+    },
+    {
+      "epoch": 0.8996535894330695,
+      "grad_norm": 0.45677739202857504,
+      "learning_rate": 1.3619008353585873e-07,
+      "loss": 0.1413,
+      "step": 3993
+    },
+    {
+      "epoch": 0.8998788971188779,
+      "grad_norm": 0.4915720070809868,
+      "learning_rate": 1.3558426848023165e-07,
+      "loss": 0.1444,
+      "step": 3994
+    },
+    {
+      "epoch": 0.9001042048046864,
+      "grad_norm": 0.4613124404503582,
+      "learning_rate": 1.3497976629652882e-07,
+      "loss": 0.1366,
+      "step": 3995
+    },
+    {
+      "epoch": 0.9003295124904949,
+      "grad_norm": 0.4866944759203155,
+      "learning_rate": 1.3437657732040783e-07,
+      "loss": 0.1401,
+      "step": 3996
+    },
+    {
+      "epoch": 0.9005548201763033,
+      "grad_norm": 0.48683156668560257,
+      "learning_rate": 1.337747018867977e-07,
+      "loss": 0.144,
+      "step": 3997
+    },
+    {
+      "epoch": 0.9007801278621117,
+      "grad_norm": 0.4795408253255835,
+      "learning_rate": 1.3317414032989668e-07,
+      "loss": 0.139,
+      "step": 3998
+    },
+    {
+      "epoch": 0.9010054355479201,
+      "grad_norm": 0.458705770255175,
+      "learning_rate": 1.3257489298317466e-07,
+      "loss": 0.1388,
+      "step": 3999
+    },
+    {
+      "epoch": 0.9012307432337285,
+      "grad_norm": 0.43108938142225783,
+      "learning_rate": 1.3197696017937106e-07,
+      "loss": 0.1278,
+      "step": 4000
+    },
+    {
+      "epoch": 0.9012307432337285,
+      "eval_loss": 0.14278066158294678,
+      "eval_runtime": 56.8248,
+      "eval_samples_per_second": 50.506,
+      "eval_steps_per_second": 6.318,
+      "step": 4000
+    },
+    {
+      "epoch": 0.901456050919537,
+      "grad_norm": 0.4664469040013792,
+      "learning_rate": 1.3138034225049583e-07,
+      "loss": 0.14,
+      "step": 4001
+    },
+    {
+      "epoch": 0.9016813586053454,
+      "grad_norm": 0.4754535001909423,
+      "learning_rate": 1.3078503952782845e-07,
+      "loss": 0.1505,
+      "step": 4002
+    },
+    {
+      "epoch": 0.9019066662911539,
+      "grad_norm": 0.4941711804629973,
+      "learning_rate": 1.3019105234191865e-07,
+      "loss": 0.1508,
+      "step": 4003
+    },
+    {
+      "epoch": 0.9021319739769623,
+      "grad_norm": 0.47736169945139195,
+      "learning_rate": 1.2959838102258537e-07,
+      "loss": 0.1426,
+      "step": 4004
+    },
+    {
+      "epoch": 0.9023572816627707,
+      "grad_norm": 0.4702410812281165,
+      "learning_rate": 1.2900702589891652e-07,
+      "loss": 0.1378,
+      "step": 4005
+    },
+    {
+      "epoch": 0.9025825893485792,
+      "grad_norm": 0.4802058798581131,
+      "learning_rate": 1.2841698729927022e-07,
+      "loss": 0.151,
+      "step": 4006
+    },
+    {
+      "epoch": 0.9028078970343876,
+      "grad_norm": 0.4718303474580779,
+      "learning_rate": 1.278282655512722e-07,
+      "loss": 0.1477,
+      "step": 4007
+    },
+    {
+      "epoch": 0.903033204720196,
+      "grad_norm": 0.44777196172649736,
+      "learning_rate": 1.272408609818182e-07,
+      "loss": 0.1367,
+      "step": 4008
+    },
+    {
+      "epoch": 0.9032585124060044,
+      "grad_norm": 0.4410043019443886,
+      "learning_rate": 1.2665477391707203e-07,
+      "loss": 0.1253,
+      "step": 4009
+    },
+    {
+      "epoch": 0.9034838200918128,
+      "grad_norm": 0.46387277123794624,
+      "learning_rate": 1.2607000468246533e-07,
+      "loss": 0.1329,
+      "step": 4010
+    },
+    {
+      "epoch": 0.9037091277776214,
+      "grad_norm": 0.4694158218919607,
+      "learning_rate": 1.2548655360269974e-07,
+      "loss": 0.1418,
+      "step": 4011
+    },
+    {
+      "epoch": 0.9039344354634298,
+      "grad_norm": 0.4774664008686541,
+      "learning_rate": 1.2490442100174278e-07,
+      "loss": 0.1419,
+      "step": 4012
+    },
+    {
+      "epoch": 0.9041597431492382,
+      "grad_norm": 0.45399983828562024,
+      "learning_rate": 1.243236072028317e-07,
+      "loss": 0.141,
+      "step": 4013
+    },
+    {
+      "epoch": 0.9043850508350466,
+      "grad_norm": 0.4685371261241214,
+      "learning_rate": 1.237441125284708e-07,
+      "loss": 0.1385,
+      "step": 4014
+    },
+    {
+      "epoch": 0.904610358520855,
+      "grad_norm": 0.4728214519874054,
+      "learning_rate": 1.2316593730043154e-07,
+      "loss": 0.1494,
+      "step": 4015
+    },
+    {
+      "epoch": 0.9048356662066634,
+      "grad_norm": 0.46208919509274266,
+      "learning_rate": 1.2258908183975322e-07,
+      "loss": 0.1321,
+      "step": 4016
+    },
+    {
+      "epoch": 0.9050609738924719,
+      "grad_norm": 0.4586787135362811,
+      "learning_rate": 1.2201354646674212e-07,
+      "loss": 0.1317,
+      "step": 4017
+    },
+    {
+      "epoch": 0.9052862815782804,
+      "grad_norm": 0.4856955819035042,
+      "learning_rate": 1.2143933150097154e-07,
+      "loss": 0.1462,
+      "step": 4018
+    },
+    {
+      "epoch": 0.9055115892640888,
+      "grad_norm": 0.4708554386279845,
+      "learning_rate": 1.2086643726128194e-07,
+      "loss": 0.1355,
+      "step": 4019
+    },
+    {
+      "epoch": 0.9057368969498972,
+      "grad_norm": 0.4975683906389299,
+      "learning_rate": 1.2029486406577972e-07,
+      "loss": 0.1393,
+      "step": 4020
+    },
+    {
+      "epoch": 0.9059622046357056,
+      "grad_norm": 0.471183068628529,
+      "learning_rate": 1.1972461223183878e-07,
+      "loss": 0.1412,
+      "step": 4021
+    },
+    {
+      "epoch": 0.9061875123215141,
+      "grad_norm": 0.4920568295014266,
+      "learning_rate": 1.191556820760978e-07,
+      "loss": 0.1517,
+      "step": 4022
+    },
+    {
+      "epoch": 0.9064128200073225,
+      "grad_norm": 0.48367463833210245,
+      "learning_rate": 1.1858807391446319e-07,
+      "loss": 0.1405,
+      "step": 4023
+    },
+    {
+      "epoch": 0.9066381276931309,
+      "grad_norm": 0.5003956217619391,
+      "learning_rate": 1.1802178806210624e-07,
+      "loss": 0.1469,
+      "step": 4024
+    },
+    {
+      "epoch": 0.9068634353789393,
+      "grad_norm": 0.4810574594091294,
+      "learning_rate": 1.1745682483346454e-07,
+      "loss": 0.1477,
+      "step": 4025
+    },
+    {
+      "epoch": 0.9070887430647478,
+      "grad_norm": 0.4531146593689317,
+      "learning_rate": 1.1689318454224191e-07,
+      "loss": 0.1296,
+      "step": 4026
+    },
+    {
+      "epoch": 0.9073140507505563,
+      "grad_norm": 0.4575572613766583,
+      "learning_rate": 1.1633086750140521e-07,
+      "loss": 0.1261,
+      "step": 4027
+    },
+    {
+      "epoch": 0.9075393584363647,
+      "grad_norm": 0.48780587471832954,
+      "learning_rate": 1.1576987402318884e-07,
+      "loss": 0.1459,
+      "step": 4028
+    },
+    {
+      "epoch": 0.9077646661221731,
+      "grad_norm": 0.46528623662642415,
+      "learning_rate": 1.1521020441909226e-07,
+      "loss": 0.1351,
+      "step": 4029
+    },
+    {
+      "epoch": 0.9079899738079815,
+      "grad_norm": 0.4966456351879077,
+      "learning_rate": 1.1465185899987797e-07,
+      "loss": 0.1479,
+      "step": 4030
+    },
+    {
+      "epoch": 0.9082152814937899,
+      "grad_norm": 0.47403911627624085,
+      "learning_rate": 1.140948380755752e-07,
+      "loss": 0.1459,
+      "step": 4031
+    },
+    {
+      "epoch": 0.9084405891795984,
+      "grad_norm": 0.48887357002467957,
+      "learning_rate": 1.1353914195547655e-07,
+      "loss": 0.1445,
+      "step": 4032
+    },
+    {
+      "epoch": 0.9086658968654068,
+      "grad_norm": 0.43742134361709295,
+      "learning_rate": 1.1298477094813965e-07,
+      "loss": 0.1282,
+      "step": 4033
+    },
+    {
+      "epoch": 0.9088912045512153,
+      "grad_norm": 0.5078126004993612,
+      "learning_rate": 1.1243172536138547e-07,
+      "loss": 0.1539,
+      "step": 4034
+    },
+    {
+      "epoch": 0.9091165122370237,
+      "grad_norm": 0.4945032415866532,
+      "learning_rate": 1.1188000550230005e-07,
+      "loss": 0.1426,
+      "step": 4035
+    },
+    {
+      "epoch": 0.9093418199228321,
+      "grad_norm": 0.496800918331347,
+      "learning_rate": 1.1132961167723305e-07,
+      "loss": 0.1532,
+      "step": 4036
+    },
+    {
+      "epoch": 0.9095671276086406,
+      "grad_norm": 0.48007976714054434,
+      "learning_rate": 1.1078054419179724e-07,
+      "loss": 0.15,
+      "step": 4037
+    },
+    {
+      "epoch": 0.909792435294449,
+      "grad_norm": 0.47821352867896594,
+      "learning_rate": 1.1023280335086956e-07,
+      "loss": 0.1441,
+      "step": 4038
+    },
+    {
+      "epoch": 0.9100177429802574,
+      "grad_norm": 0.503025431983095,
+      "learning_rate": 1.0968638945858978e-07,
+      "loss": 0.1452,
+      "step": 4039
+    },
+    {
+      "epoch": 0.9102430506660658,
+      "grad_norm": 0.4960063821194838,
+      "learning_rate": 1.091413028183616e-07,
+      "loss": 0.1573,
+      "step": 4040
+    },
+    {
+      "epoch": 0.9104683583518742,
+      "grad_norm": 0.4669980833464851,
+      "learning_rate": 1.0859754373285125e-07,
+      "loss": 0.1506,
+      "step": 4041
+    },
+    {
+      "epoch": 0.9106936660376828,
+      "grad_norm": 0.46358731915184515,
+      "learning_rate": 1.0805511250398748e-07,
+      "loss": 0.137,
+      "step": 4042
+    },
+    {
+      "epoch": 0.9109189737234912,
+      "grad_norm": 0.46194698650882016,
+      "learning_rate": 1.0751400943296269e-07,
+      "loss": 0.1361,
+      "step": 4043
+    },
+    {
+      "epoch": 0.9111442814092996,
+      "grad_norm": 0.4797787938586833,
+      "learning_rate": 1.06974234820231e-07,
+      "loss": 0.1326,
+      "step": 4044
+    },
+    {
+      "epoch": 0.911369589095108,
+      "grad_norm": 0.45647525029825825,
+      "learning_rate": 1.0643578896550877e-07,
+      "loss": 0.1399,
+      "step": 4045
+    },
+    {
+      "epoch": 0.9115948967809164,
+      "grad_norm": 0.4769578133095442,
+      "learning_rate": 1.0589867216777544e-07,
+      "loss": 0.1483,
+      "step": 4046
+    },
+    {
+      "epoch": 0.9118202044667248,
+      "grad_norm": 0.47151532859509643,
+      "learning_rate": 1.0536288472527162e-07,
+      "loss": 0.14,
+      "step": 4047
+    },
+    {
+      "epoch": 0.9120455121525333,
+      "grad_norm": 0.47164880427720157,
+      "learning_rate": 1.0482842693550044e-07,
+      "loss": 0.1369,
+      "step": 4048
+    },
+    {
+      "epoch": 0.9122708198383417,
+      "grad_norm": 0.450133586818689,
+      "learning_rate": 1.042952990952259e-07,
+      "loss": 0.1373,
+      "step": 4049
+    },
+    {
+      "epoch": 0.9124961275241502,
+      "grad_norm": 0.5002594849950737,
+      "learning_rate": 1.0376350150047427e-07,
+      "loss": 0.1487,
+      "step": 4050
+    },
+    {
+      "epoch": 0.9127214352099586,
+      "grad_norm": 0.45280315659636944,
+      "learning_rate": 1.032330344465332e-07,
+      "loss": 0.1336,
+      "step": 4051
+    },
+    {
+      "epoch": 0.912946742895767,
+      "grad_norm": 0.49636257095046055,
+      "learning_rate": 1.0270389822795073e-07,
+      "loss": 0.1394,
+      "step": 4052
+    },
+    {
+      "epoch": 0.9131720505815755,
+      "grad_norm": 0.4662242703415579,
+      "learning_rate": 1.0217609313853738e-07,
+      "loss": 0.1396,
+      "step": 4053
+    },
+    {
+      "epoch": 0.9133973582673839,
+      "grad_norm": 0.4626581189870639,
+      "learning_rate": 1.0164961947136232e-07,
+      "loss": 0.1399,
+      "step": 4054
+    },
+    {
+      "epoch": 0.9136226659531923,
+      "grad_norm": 0.4780719271605745,
+      "learning_rate": 1.0112447751875809e-07,
+      "loss": 0.1474,
+      "step": 4055
+    },
+    {
+      "epoch": 0.9138479736390007,
+      "grad_norm": 0.4927695082239918,
+      "learning_rate": 1.0060066757231535e-07,
+      "loss": 0.1347,
+      "step": 4056
+    },
+    {
+      "epoch": 0.9140732813248091,
+      "grad_norm": 0.4538682665106712,
+      "learning_rate": 1.0007818992288671e-07,
+      "loss": 0.1287,
+      "step": 4057
+    },
+    {
+      "epoch": 0.9142985890106177,
+      "grad_norm": 0.45836132297611115,
+      "learning_rate": 9.955704486058482e-08,
+      "loss": 0.1377,
+      "step": 4058
+    },
+    {
+      "epoch": 0.9145238966964261,
+      "grad_norm": 0.46646988612796003,
+      "learning_rate": 9.903723267478154e-08,
+      "loss": 0.1278,
+      "step": 4059
+    },
+    {
+      "epoch": 0.9147492043822345,
+      "grad_norm": 0.4735433165512558,
+      "learning_rate": 9.85187536541099e-08,
+      "loss": 0.1333,
+      "step": 4060
+    },
+    {
+      "epoch": 0.9149745120680429,
+      "grad_norm": 0.46501747980793207,
+      "learning_rate": 9.800160808646154e-08,
+      "loss": 0.1339,
+      "step": 4061
+    },
+    {
+      "epoch": 0.9151998197538513,
+      "grad_norm": 0.4735027259975002,
+      "learning_rate": 9.748579625898758e-08,
+      "loss": 0.1399,
+      "step": 4062
+    },
+    {
+      "epoch": 0.9154251274396598,
+      "grad_norm": 0.5016523883899949,
+      "learning_rate": 9.697131845810032e-08,
+      "loss": 0.148,
+      "step": 4063
+    },
+    {
+      "epoch": 0.9156504351254682,
+      "grad_norm": 0.4585294421782351,
+      "learning_rate": 9.645817496946902e-08,
+      "loss": 0.1382,
+      "step": 4064
+    },
+    {
+      "epoch": 0.9158757428112767,
+      "grad_norm": 0.49296755886219573,
+      "learning_rate": 9.594636607802355e-08,
+      "loss": 0.1529,
+      "step": 4065
+    },
+    {
+      "epoch": 0.9161010504970851,
+      "grad_norm": 0.4840871686662579,
+      "learning_rate": 9.54358920679524e-08,
+      "loss": 0.1504,
+      "step": 4066
+    },
+    {
+      "epoch": 0.9163263581828935,
+      "grad_norm": 0.4673411493667281,
+      "learning_rate": 9.492675322270273e-08,
+      "loss": 0.1364,
+      "step": 4067
+    },
+    {
+      "epoch": 0.916551665868702,
+      "grad_norm": 0.45852533144391544,
+      "learning_rate": 9.441894982498035e-08,
+      "loss": 0.1248,
+      "step": 4068
+    },
+    {
+      "epoch": 0.9167769735545104,
+      "grad_norm": 0.5194322131946961,
+      "learning_rate": 9.391248215674942e-08,
+      "loss": 0.1576,
+      "step": 4069
+    },
+    {
+      "epoch": 0.9170022812403188,
+      "grad_norm": 0.4841221144035149,
+      "learning_rate": 9.340735049923277e-08,
+      "loss": 0.1455,
+      "step": 4070
+    },
+    {
+      "epoch": 0.9172275889261272,
+      "grad_norm": 0.43263175772768625,
+      "learning_rate": 9.290355513291105e-08,
+      "loss": 0.1208,
+      "step": 4071
+    },
+    {
+      "epoch": 0.9174528966119356,
+      "grad_norm": 0.4527159715269973,
+      "learning_rate": 9.24010963375227e-08,
+      "loss": 0.1368,
+      "step": 4072
+    },
+    {
+      "epoch": 0.9176782042977442,
+      "grad_norm": 0.5077271246523941,
+      "learning_rate": 9.189997439206538e-08,
+      "loss": 0.164,
+      "step": 4073
+    },
+    {
+      "epoch": 0.9179035119835526,
+      "grad_norm": 0.46080076191265873,
+      "learning_rate": 9.140018957479236e-08,
+      "loss": 0.1294,
+      "step": 4074
+    },
+    {
+      "epoch": 0.918128819669361,
+      "grad_norm": 0.46911421219992144,
+      "learning_rate": 9.090174216321607e-08,
+      "loss": 0.1381,
+      "step": 4075
+    },
+    {
+      "epoch": 0.9183541273551694,
+      "grad_norm": 0.504179840271321,
+      "learning_rate": 9.040463243410541e-08,
+      "loss": 0.1602,
+      "step": 4076
+    },
+    {
+      "epoch": 0.9185794350409778,
+      "grad_norm": 0.4794307845268256,
+      "learning_rate": 8.990886066348764e-08,
+      "loss": 0.1312,
+      "step": 4077
+    },
+    {
+      "epoch": 0.9188047427267862,
+      "grad_norm": 0.49026049223089996,
+      "learning_rate": 8.941442712664561e-08,
+      "loss": 0.1452,
+      "step": 4078
+    },
+    {
+      "epoch": 0.9190300504125947,
+      "grad_norm": 0.49867226746965787,
+      "learning_rate": 8.892133209811971e-08,
+      "loss": 0.1459,
+      "step": 4079
+    },
+    {
+      "epoch": 0.9192553580984031,
+      "grad_norm": 0.46105305653005,
+      "learning_rate": 8.842957585170814e-08,
+      "loss": 0.1409,
+      "step": 4080
+    },
+    {
+      "epoch": 0.9194806657842116,
+      "grad_norm": 0.45755223484497676,
+      "learning_rate": 8.79391586604636e-08,
+      "loss": 0.1317,
+      "step": 4081
+    },
+    {
+      "epoch": 0.91970597347002,
+      "grad_norm": 0.4910936827240193,
+      "learning_rate": 8.745008079669742e-08,
+      "loss": 0.153,
+      "step": 4082
+    },
+    {
+      "epoch": 0.9199312811558285,
+      "grad_norm": 0.4776348509081474,
+      "learning_rate": 8.696234253197599e-08,
+      "loss": 0.1509,
+      "step": 4083
+    },
+    {
+      "epoch": 0.9201565888416369,
+      "grad_norm": 0.4892957646542491,
+      "learning_rate": 8.647594413712212e-08,
+      "loss": 0.1449,
+      "step": 4084
+    },
+    {
+      "epoch": 0.9203818965274453,
+      "grad_norm": 0.46500853465974756,
+      "learning_rate": 8.599088588221504e-08,
+      "loss": 0.1334,
+      "step": 4085
+    },
+    {
+      "epoch": 0.9206072042132537,
+      "grad_norm": 0.4729412778091547,
+      "learning_rate": 8.550716803658904e-08,
+      "loss": 0.1425,
+      "step": 4086
+    },
+    {
+      "epoch": 0.9208325118990621,
+      "grad_norm": 0.4941003039455095,
+      "learning_rate": 8.502479086883481e-08,
+      "loss": 0.1575,
+      "step": 4087
+    },
+    {
+      "epoch": 0.9210578195848705,
+      "grad_norm": 0.4798745839790279,
+      "learning_rate": 8.454375464679865e-08,
+      "loss": 0.1542,
+      "step": 4088
+    },
+    {
+      "epoch": 0.9212831272706791,
+      "grad_norm": 0.4965004585488724,
+      "learning_rate": 8.406405963758162e-08,
+      "loss": 0.1532,
+      "step": 4089
+    },
+    {
+      "epoch": 0.9215084349564875,
+      "grad_norm": 0.46374661519707955,
+      "learning_rate": 8.358570610754097e-08,
+      "loss": 0.1438,
+      "step": 4090
+    },
+    {
+      "epoch": 0.9217337426422959,
+      "grad_norm": 0.4948131528817748,
+      "learning_rate": 8.310869432228808e-08,
+      "loss": 0.1488,
+      "step": 4091
+    },
+    {
+      "epoch": 0.9219590503281043,
+      "grad_norm": 0.453044774452747,
+      "learning_rate": 8.263302454669025e-08,
+      "loss": 0.1318,
+      "step": 4092
+    },
+    {
+      "epoch": 0.9221843580139127,
+      "grad_norm": 0.48266271674169603,
+      "learning_rate": 8.215869704486873e-08,
+      "loss": 0.1493,
+      "step": 4093
+    },
+    {
+      "epoch": 0.9224096656997212,
+      "grad_norm": 0.4585831905667495,
+      "learning_rate": 8.168571208020032e-08,
+      "loss": 0.1372,
+      "step": 4094
+    },
+    {
+      "epoch": 0.9226349733855296,
+      "grad_norm": 0.4866299635039302,
+      "learning_rate": 8.121406991531577e-08,
+      "loss": 0.1455,
+      "step": 4095
+    },
+    {
+      "epoch": 0.922860281071338,
+      "grad_norm": 0.4478029123064623,
+      "learning_rate": 8.074377081210033e-08,
+      "loss": 0.1312,
+      "step": 4096
+    },
+    {
+      "epoch": 0.9230855887571465,
+      "grad_norm": 0.4665798318007941,
+      "learning_rate": 8.027481503169371e-08,
+      "loss": 0.1432,
+      "step": 4097
+    },
+    {
+      "epoch": 0.9233108964429549,
+      "grad_norm": 0.4529312789736691,
+      "learning_rate": 7.980720283448957e-08,
+      "loss": 0.1324,
+      "step": 4098
+    },
+    {
+      "epoch": 0.9235362041287634,
+      "grad_norm": 0.49127056991795576,
+      "learning_rate": 7.934093448013492e-08,
+      "loss": 0.1518,
+      "step": 4099
+    },
+    {
+      "epoch": 0.9237615118145718,
+      "grad_norm": 0.47449397767122997,
+      "learning_rate": 7.887601022753238e-08,
+      "loss": 0.1344,
+      "step": 4100
+    },
+    {
+      "epoch": 0.9239868195003802,
+      "grad_norm": 0.45542770853975467,
+      "learning_rate": 7.841243033483575e-08,
+      "loss": 0.1261,
+      "step": 4101
+    },
+    {
+      "epoch": 0.9242121271861886,
+      "grad_norm": 0.4490597747023126,
+      "learning_rate": 7.795019505945495e-08,
+      "loss": 0.138,
+      "step": 4102
+    },
+    {
+      "epoch": 0.924437434871997,
+      "grad_norm": 0.4777794405361795,
+      "learning_rate": 7.748930465805105e-08,
+      "loss": 0.1442,
+      "step": 4103
+    },
+    {
+      "epoch": 0.9246627425578055,
+      "grad_norm": 0.4782645372300299,
+      "learning_rate": 7.702975938653934e-08,
+      "loss": 0.1453,
+      "step": 4104
+    },
+    {
+      "epoch": 0.924888050243614,
+      "grad_norm": 0.4797534525418542,
+      "learning_rate": 7.657155950008904e-08,
+      "loss": 0.1422,
+      "step": 4105
+    },
+    {
+      "epoch": 0.9251133579294224,
+      "grad_norm": 0.4859477729123614,
+      "learning_rate": 7.611470525312054e-08,
+      "loss": 0.1422,
+      "step": 4106
+    },
+    {
+      "epoch": 0.9253386656152308,
+      "grad_norm": 0.47914845153238234,
+      "learning_rate": 7.565919689930839e-08,
+      "loss": 0.1455,
+      "step": 4107
+    },
+    {
+      "epoch": 0.9255639733010392,
+      "grad_norm": 0.462832667536284,
+      "learning_rate": 7.520503469157947e-08,
+      "loss": 0.1348,
+      "step": 4108
+    },
+    {
+      "epoch": 0.9257892809868477,
+      "grad_norm": 0.47049952129817785,
+      "learning_rate": 7.47522188821126e-08,
+      "loss": 0.1407,
+      "step": 4109
+    },
+    {
+      "epoch": 0.9260145886726561,
+      "grad_norm": 0.44762021394675944,
+      "learning_rate": 7.430074972234053e-08,
+      "loss": 0.1318,
+      "step": 4110
+    },
+    {
+      "epoch": 0.9262398963584645,
+      "grad_norm": 0.45952545767286157,
+      "learning_rate": 7.385062746294608e-08,
+      "loss": 0.134,
+      "step": 4111
+    },
+    {
+      "epoch": 0.9264652040442729,
+      "grad_norm": 0.43848643342546195,
+      "learning_rate": 7.340185235386627e-08,
+      "loss": 0.1266,
+      "step": 4112
+    },
+    {
+      "epoch": 0.9266905117300814,
+      "grad_norm": 0.48233284878708355,
+      "learning_rate": 7.29544246442887e-08,
+      "loss": 0.1438,
+      "step": 4113
+    },
+    {
+      "epoch": 0.9269158194158899,
+      "grad_norm": 0.48040723283418874,
+      "learning_rate": 7.250834458265355e-08,
+      "loss": 0.1354,
+      "step": 4114
+    },
+    {
+      "epoch": 0.9271411271016983,
+      "grad_norm": 0.47447604347591066,
+      "learning_rate": 7.206361241665266e-08,
+      "loss": 0.1392,
+      "step": 4115
+    },
+    {
+      "epoch": 0.9273664347875067,
+      "grad_norm": 0.46371914947128257,
+      "learning_rate": 7.162022839322824e-08,
+      "loss": 0.132,
+      "step": 4116
+    },
+    {
+      "epoch": 0.9275917424733151,
+      "grad_norm": 0.47355152598312406,
+      "learning_rate": 7.117819275857613e-08,
+      "loss": 0.136,
+      "step": 4117
+    },
+    {
+      "epoch": 0.9278170501591235,
+      "grad_norm": 0.4928462672275944,
+      "learning_rate": 7.073750575814136e-08,
+      "loss": 0.1492,
+      "step": 4118
+    },
+    {
+      "epoch": 0.9280423578449319,
+      "grad_norm": 0.47060721385062093,
+      "learning_rate": 7.029816763662129e-08,
+      "loss": 0.1392,
+      "step": 4119
+    },
+    {
+      "epoch": 0.9282676655307405,
+      "grad_norm": 0.47917164908363497,
+      "learning_rate": 6.986017863796435e-08,
+      "loss": 0.1307,
+      "step": 4120
+    },
+    {
+      "epoch": 0.9284929732165489,
+      "grad_norm": 0.48465172787325095,
+      "learning_rate": 6.94235390053688e-08,
+      "loss": 0.139,
+      "step": 4121
+    },
+    {
+      "epoch": 0.9287182809023573,
+      "grad_norm": 0.481611750105736,
+      "learning_rate": 6.898824898128515e-08,
+      "loss": 0.1363,
+      "step": 4122
+    },
+    {
+      "epoch": 0.9289435885881657,
+      "grad_norm": 0.4807778024489483,
+      "learning_rate": 6.85543088074131e-08,
+      "loss": 0.1356,
+      "step": 4123
+    },
+    {
+      "epoch": 0.9291688962739741,
+      "grad_norm": 0.49271889080241804,
+      "learning_rate": 6.81217187247038e-08,
+      "loss": 0.1464,
+      "step": 4124
+    },
+    {
+      "epoch": 0.9293942039597826,
+      "grad_norm": 0.49943470735311984,
+      "learning_rate": 6.769047897335818e-08,
+      "loss": 0.1563,
+      "step": 4125
+    },
+    {
+      "epoch": 0.929619511645591,
+      "grad_norm": 0.4562498902244851,
+      "learning_rate": 6.726058979282774e-08,
+      "loss": 0.1277,
+      "step": 4126
+    },
+    {
+      "epoch": 0.9298448193313994,
+      "grad_norm": 0.47040935352758834,
+      "learning_rate": 6.683205142181404e-08,
+      "loss": 0.1345,
+      "step": 4127
+    },
+    {
+      "epoch": 0.9300701270172079,
+      "grad_norm": 0.48800656067496256,
+      "learning_rate": 6.640486409826785e-08,
+      "loss": 0.1426,
+      "step": 4128
+    },
+    {
+      "epoch": 0.9302954347030163,
+      "grad_norm": 0.4473344699309852,
+      "learning_rate": 6.597902805939138e-08,
+      "loss": 0.1365,
+      "step": 4129
+    },
+    {
+      "epoch": 0.9305207423888248,
+      "grad_norm": 0.4979933517833155,
+      "learning_rate": 6.555454354163437e-08,
+      "loss": 0.1484,
+      "step": 4130
+    },
+    {
+      "epoch": 0.9307460500746332,
+      "grad_norm": 0.4517633354010443,
+      "learning_rate": 6.513141078069828e-08,
+      "loss": 0.1293,
+      "step": 4131
+    },
+    {
+      "epoch": 0.9309713577604416,
+      "grad_norm": 0.4946926565291331,
+      "learning_rate": 6.470963001153268e-08,
+      "loss": 0.1537,
+      "step": 4132
+    },
+    {
+      "epoch": 0.93119666544625,
+      "grad_norm": 0.48974509519757004,
+      "learning_rate": 6.428920146833606e-08,
+      "loss": 0.1462,
+      "step": 4133
+    },
+    {
+      "epoch": 0.9314219731320584,
+      "grad_norm": 0.4761209450780214,
+      "learning_rate": 6.387012538455723e-08,
+      "loss": 0.1517,
+      "step": 4134
+    },
+    {
+      "epoch": 0.9316472808178669,
+      "grad_norm": 0.4706955427126162,
+      "learning_rate": 6.345240199289365e-08,
+      "loss": 0.1438,
+      "step": 4135
+    },
+    {
+      "epoch": 0.9318725885036754,
+      "grad_norm": 0.46452407200738693,
+      "learning_rate": 6.303603152529119e-08,
+      "loss": 0.1408,
+      "step": 4136
+    },
+    {
+      "epoch": 0.9320978961894838,
+      "grad_norm": 0.46541018057868117,
+      "learning_rate": 6.262101421294547e-08,
+      "loss": 0.1343,
+      "step": 4137
+    },
+    {
+      "epoch": 0.9323232038752922,
+      "grad_norm": 0.47098480181557906,
+      "learning_rate": 6.220735028629937e-08,
+      "loss": 0.1327,
+      "step": 4138
+    },
+    {
+      "epoch": 0.9325485115611006,
+      "grad_norm": 0.4755568983527909,
+      "learning_rate": 6.179503997504554e-08,
+      "loss": 0.1395,
+      "step": 4139
+    },
+    {
+      "epoch": 0.932773819246909,
+      "grad_norm": 0.4815905104832014,
+      "learning_rate": 6.13840835081242e-08,
+      "loss": 0.1365,
+      "step": 4140
+    },
+    {
+      "epoch": 0.9329991269327175,
+      "grad_norm": 0.4838954462320816,
+      "learning_rate": 6.097448111372446e-08,
+      "loss": 0.15,
+      "step": 4141
+    },
+    {
+      "epoch": 0.9332244346185259,
+      "grad_norm": 0.4857125082783464,
+      "learning_rate": 6.056623301928327e-08,
+      "loss": 0.149,
+      "step": 4142
+    },
+    {
+      "epoch": 0.9334497423043343,
+      "grad_norm": 0.45053349895739203,
+      "learning_rate": 6.015933945148517e-08,
+      "loss": 0.1308,
+      "step": 4143
+    },
+    {
+      "epoch": 0.9336750499901428,
+      "grad_norm": 0.48660458250447647,
+      "learning_rate": 5.975380063626356e-08,
+      "loss": 0.1425,
+      "step": 4144
+    },
+    {
+      "epoch": 0.9339003576759513,
+      "grad_norm": 0.47940782085921785,
+      "learning_rate": 5.934961679879836e-08,
+      "loss": 0.1411,
+      "step": 4145
+    },
+    {
+      "epoch": 0.9341256653617597,
+      "grad_norm": 0.44987270917047917,
+      "learning_rate": 5.894678816351862e-08,
+      "loss": 0.133,
+      "step": 4146
+    },
+    {
+      "epoch": 0.9343509730475681,
+      "grad_norm": 0.48414334794038866,
+      "learning_rate": 5.854531495409932e-08,
+      "loss": 0.145,
+      "step": 4147
+    },
+    {
+      "epoch": 0.9345762807333765,
+      "grad_norm": 0.4639549974078448,
+      "learning_rate": 5.8145197393463806e-08,
+      "loss": 0.14,
+      "step": 4148
+    },
+    {
+      "epoch": 0.9348015884191849,
+      "grad_norm": 0.45948529746860883,
+      "learning_rate": 5.774643570378296e-08,
+      "loss": 0.1287,
+      "step": 4149
+    },
+    {
+      "epoch": 0.9350268961049933,
+      "grad_norm": 0.47429997344838176,
+      "learning_rate": 5.73490301064733e-08,
+      "loss": 0.1367,
+      "step": 4150
+    },
+    {
+      "epoch": 0.9352522037908018,
+      "grad_norm": 0.48066514481237554,
+      "learning_rate": 5.695298082219997e-08,
+      "loss": 0.1521,
+      "step": 4151
+    },
+    {
+      "epoch": 0.9354775114766103,
+      "grad_norm": 0.46185271302563835,
+      "learning_rate": 5.6558288070874544e-08,
+      "loss": 0.1349,
+      "step": 4152
+    },
+    {
+      "epoch": 0.9357028191624187,
+      "grad_norm": 0.4992867840736411,
+      "learning_rate": 5.616495207165451e-08,
+      "loss": 0.1562,
+      "step": 4153
+    },
+    {
+      "epoch": 0.9359281268482271,
+      "grad_norm": 0.48381812287783943,
+      "learning_rate": 5.577297304294543e-08,
+      "loss": 0.142,
+      "step": 4154
+    },
+    {
+      "epoch": 0.9361534345340355,
+      "grad_norm": 0.46829283932499904,
+      "learning_rate": 5.538235120239821e-08,
+      "loss": 0.1339,
+      "step": 4155
+    },
+    {
+      "epoch": 0.936378742219844,
+      "grad_norm": 0.4734966918946892,
+      "learning_rate": 5.4993086766910733e-08,
+      "loss": 0.1383,
+      "step": 4156
+    },
+    {
+      "epoch": 0.9366040499056524,
+      "grad_norm": 0.4877136325551203,
+      "learning_rate": 5.460517995262704e-08,
+      "loss": 0.1456,
+      "step": 4157
+    },
+    {
+      "epoch": 0.9368293575914608,
+      "grad_norm": 0.47711868327825846,
+      "learning_rate": 5.421863097493707e-08,
+      "loss": 0.1433,
+      "step": 4158
+    },
+    {
+      "epoch": 0.9370546652772692,
+      "grad_norm": 0.47882405572711156,
+      "learning_rate": 5.383344004847774e-08,
+      "loss": 0.147,
+      "step": 4159
+    },
+    {
+      "epoch": 0.9372799729630777,
+      "grad_norm": 0.4542593431360111,
+      "learning_rate": 5.344960738713018e-08,
+      "loss": 0.1303,
+      "step": 4160
+    },
+    {
+      "epoch": 0.9375052806488862,
+      "grad_norm": 0.49811599304780374,
+      "learning_rate": 5.3067133204023344e-08,
+      "loss": 0.1563,
+      "step": 4161
+    },
+    {
+      "epoch": 0.9377305883346946,
+      "grad_norm": 0.45144422788710814,
+      "learning_rate": 5.268601771153042e-08,
+      "loss": 0.1374,
+      "step": 4162
+    },
+    {
+      "epoch": 0.937955896020503,
+      "grad_norm": 0.4755448216457163,
+      "learning_rate": 5.230626112127046e-08,
+      "loss": 0.1341,
+      "step": 4163
+    },
+    {
+      "epoch": 0.9381812037063114,
+      "grad_norm": 0.4666514696365624,
+      "learning_rate": 5.192786364410868e-08,
+      "loss": 0.139,
+      "step": 4164
+    },
+    {
+      "epoch": 0.9384065113921198,
+      "grad_norm": 0.44495670514922453,
+      "learning_rate": 5.15508254901545e-08,
+      "loss": 0.132,
+      "step": 4165
+    },
+    {
+      "epoch": 0.9386318190779283,
+      "grad_norm": 0.494033372148409,
+      "learning_rate": 5.117514686876379e-08,
+      "loss": 0.1376,
+      "step": 4166
+    },
+    {
+      "epoch": 0.9388571267637368,
+      "grad_norm": 0.47118120206769226,
+      "learning_rate": 5.080082798853664e-08,
+      "loss": 0.1416,
+      "step": 4167
+    },
+    {
+      "epoch": 0.9390824344495452,
+      "grad_norm": 0.47736361166036123,
+      "learning_rate": 5.0427869057317894e-08,
+      "loss": 0.1435,
+      "step": 4168
+    },
+    {
+      "epoch": 0.9393077421353536,
+      "grad_norm": 0.47639039780312875,
+      "learning_rate": 5.0056270282198286e-08,
+      "loss": 0.1429,
+      "step": 4169
+    },
+    {
+      "epoch": 0.939533049821162,
+      "grad_norm": 0.4860936697164675,
+      "learning_rate": 4.9686031869512486e-08,
+      "loss": 0.1416,
+      "step": 4170
+    },
+    {
+      "epoch": 0.9397583575069705,
+      "grad_norm": 0.482815171118602,
+      "learning_rate": 4.93171540248405e-08,
+      "loss": 0.1353,
+      "step": 4171
+    },
+    {
+      "epoch": 0.9399836651927789,
+      "grad_norm": 0.5103137344503779,
+      "learning_rate": 4.89496369530057e-08,
+      "loss": 0.155,
+      "step": 4172
+    },
+    {
+      "epoch": 0.9402089728785873,
+      "grad_norm": 0.4838258789983965,
+      "learning_rate": 4.858348085807735e-08,
+      "loss": 0.1422,
+      "step": 4173
+    },
+    {
+      "epoch": 0.9404342805643957,
+      "grad_norm": 0.4851697238429523,
+      "learning_rate": 4.8218685943368094e-08,
+      "loss": 0.1495,
+      "step": 4174
+    },
+    {
+      "epoch": 0.9406595882502042,
+      "grad_norm": 0.49732599987854065,
+      "learning_rate": 4.7855252411434516e-08,
+      "loss": 0.1492,
+      "step": 4175
+    },
+    {
+      "epoch": 0.9408848959360127,
+      "grad_norm": 0.4717565593998016,
+      "learning_rate": 4.7493180464078246e-08,
+      "loss": 0.1388,
+      "step": 4176
+    },
+    {
+      "epoch": 0.9411102036218211,
+      "grad_norm": 0.4664459236488136,
+      "learning_rate": 4.713247030234402e-08,
+      "loss": 0.1425,
+      "step": 4177
+    },
+    {
+      "epoch": 0.9413355113076295,
+      "grad_norm": 0.4811319686765736,
+      "learning_rate": 4.677312212652108e-08,
+      "loss": 0.1385,
+      "step": 4178
+    },
+    {
+      "epoch": 0.9415608189934379,
+      "grad_norm": 0.48387282838605733,
+      "learning_rate": 4.641513613614174e-08,
+      "loss": 0.1427,
+      "step": 4179
+    },
+    {
+      "epoch": 0.9417861266792463,
+      "grad_norm": 0.48546839650306356,
+      "learning_rate": 4.605851252998256e-08,
+      "loss": 0.1429,
+      "step": 4180
+    },
+    {
+      "epoch": 0.9420114343650547,
+      "grad_norm": 0.49348764821190694,
+      "learning_rate": 4.570325150606292e-08,
+      "loss": 0.1404,
+      "step": 4181
+    },
+    {
+      "epoch": 0.9422367420508632,
+      "grad_norm": 0.47553158908716175,
+      "learning_rate": 4.5349353261646414e-08,
+      "loss": 0.1402,
+      "step": 4182
+    },
+    {
+      "epoch": 0.9424620497366717,
+      "grad_norm": 0.43968316669747853,
+      "learning_rate": 4.4996817993239464e-08,
+      "loss": 0.122,
+      "step": 4183
+    },
+    {
+      "epoch": 0.9426873574224801,
+      "grad_norm": 0.48967752450494284,
+      "learning_rate": 4.464564589659187e-08,
+      "loss": 0.1517,
+      "step": 4184
+    },
+    {
+      "epoch": 0.9429126651082885,
+      "grad_norm": 0.484849385831537,
+      "learning_rate": 4.4295837166696e-08,
+      "loss": 0.1415,
+      "step": 4185
+    },
+    {
+      "epoch": 0.943137972794097,
+      "grad_norm": 0.48741048696615075,
+      "learning_rate": 4.3947391997787857e-08,
+      "loss": 0.1493,
+      "step": 4186
+    },
+    {
+      "epoch": 0.9433632804799054,
+      "grad_norm": 0.48746792079410006,
+      "learning_rate": 4.360031058334602e-08,
+      "loss": 0.1549,
+      "step": 4187
+    },
+    {
+      "epoch": 0.9435885881657138,
+      "grad_norm": 0.47040559458568837,
+      "learning_rate": 4.325459311609187e-08,
+      "loss": 0.1443,
+      "step": 4188
+    },
+    {
+      "epoch": 0.9438138958515222,
+      "grad_norm": 0.4896748350854861,
+      "learning_rate": 4.291023978798964e-08,
+      "loss": 0.1492,
+      "step": 4189
+    },
+    {
+      "epoch": 0.9440392035373306,
+      "grad_norm": 0.5119548403533473,
+      "learning_rate": 4.256725079024554e-08,
+      "loss": 0.1504,
+      "step": 4190
+    },
+    {
+      "epoch": 0.9442645112231391,
+      "grad_norm": 0.48365987538100963,
+      "learning_rate": 4.22256263133089e-08,
+      "loss": 0.1449,
+      "step": 4191
+    },
+    {
+      "epoch": 0.9444898189089476,
+      "grad_norm": 0.4737091419191897,
+      "learning_rate": 4.1885366546870754e-08,
+      "loss": 0.145,
+      "step": 4192
+    },
+    {
+      "epoch": 0.944715126594756,
+      "grad_norm": 0.49621450405580125,
+      "learning_rate": 4.1546471679864975e-08,
+      "loss": 0.1494,
+      "step": 4193
+    },
+    {
+      "epoch": 0.9449404342805644,
+      "grad_norm": 0.4610915608131055,
+      "learning_rate": 4.120894190046687e-08,
+      "loss": 0.1307,
+      "step": 4194
+    },
+    {
+      "epoch": 0.9451657419663728,
+      "grad_norm": 0.4546180043565135,
+      "learning_rate": 4.087277739609458e-08,
+      "loss": 0.1289,
+      "step": 4195
+    },
+    {
+      "epoch": 0.9453910496521812,
+      "grad_norm": 0.4732411285427456,
+      "learning_rate": 4.053797835340739e-08,
+      "loss": 0.1425,
+      "step": 4196
+    },
+    {
+      "epoch": 0.9456163573379897,
+      "grad_norm": 0.4353493843568312,
+      "learning_rate": 4.020454495830689e-08,
+      "loss": 0.1247,
+      "step": 4197
+    },
+    {
+      "epoch": 0.9458416650237981,
+      "grad_norm": 0.4696346129690689,
+      "learning_rate": 3.987247739593636e-08,
+      "loss": 0.1456,
+      "step": 4198
+    },
+    {
+      "epoch": 0.9460669727096066,
+      "grad_norm": 0.4470804636692044,
+      "learning_rate": 3.9541775850679975e-08,
+      "loss": 0.1329,
+      "step": 4199
+    },
+    {
+      "epoch": 0.946292280395415,
+      "grad_norm": 0.45593946874537,
+      "learning_rate": 3.9212440506164465e-08,
+      "loss": 0.1372,
+      "step": 4200
+    },
+    {
+      "epoch": 0.9465175880812234,
+      "grad_norm": 0.4724120173559638,
+      "learning_rate": 3.888447154525771e-08,
+      "loss": 0.1496,
+      "step": 4201
+    },
+    {
+      "epoch": 0.9467428957670319,
+      "grad_norm": 0.4618892258266053,
+      "learning_rate": 3.855786915006793e-08,
+      "loss": 0.134,
+      "step": 4202
+    },
+    {
+      "epoch": 0.9469682034528403,
+      "grad_norm": 0.44804853648087367,
+      "learning_rate": 3.8232633501945896e-08,
+      "loss": 0.1244,
+      "step": 4203
+    },
+    {
+      "epoch": 0.9471935111386487,
+      "grad_norm": 0.4958248010923297,
+      "learning_rate": 3.790876478148242e-08,
+      "loss": 0.148,
+      "step": 4204
+    },
+    {
+      "epoch": 0.9474188188244571,
+      "grad_norm": 0.4892013513348879,
+      "learning_rate": 3.758626316850977e-08,
+      "loss": 0.1513,
+      "step": 4205
+    },
+    {
+      "epoch": 0.9476441265102655,
+      "grad_norm": 0.5008552166510983,
+      "learning_rate": 3.726512884210165e-08,
+      "loss": 0.1474,
+      "step": 4206
+    },
+    {
+      "epoch": 0.9478694341960741,
+      "grad_norm": 0.46831375179856266,
+      "learning_rate": 3.694536198057097e-08,
+      "loss": 0.1344,
+      "step": 4207
+    },
+    {
+      "epoch": 0.9480947418818825,
+      "grad_norm": 0.47748164992826336,
+      "learning_rate": 3.6626962761473205e-08,
+      "loss": 0.1448,
+      "step": 4208
+    },
+    {
+      "epoch": 0.9483200495676909,
+      "grad_norm": 0.4695762282250701,
+      "learning_rate": 3.630993136160332e-08,
+      "loss": 0.1285,
+      "step": 4209
+    },
+    {
+      "epoch": 0.9485453572534993,
+      "grad_norm": 0.49116182161803307,
+      "learning_rate": 3.599426795699662e-08,
+      "loss": 0.1493,
+      "step": 4210
+    },
+    {
+      "epoch": 0.9487706649393077,
+      "grad_norm": 0.47400771480393467,
+      "learning_rate": 3.567997272293011e-08,
+      "loss": 0.1499,
+      "step": 4211
+    },
+    {
+      "epoch": 0.9489959726251161,
+      "grad_norm": 0.4669982124202751,
+      "learning_rate": 3.53670458339192e-08,
+      "loss": 0.1411,
+      "step": 4212
+    },
+    {
+      "epoch": 0.9492212803109246,
+      "grad_norm": 0.4546208294463335,
+      "learning_rate": 3.505548746372128e-08,
+      "loss": 0.1337,
+      "step": 4213
+    },
+    {
+      "epoch": 0.9494465879967331,
+      "grad_norm": 0.4781864888823631,
+      "learning_rate": 3.474529778533298e-08,
+      "loss": 0.1428,
+      "step": 4214
+    },
+    {
+      "epoch": 0.9496718956825415,
+      "grad_norm": 0.4628342120450911,
+      "learning_rate": 3.443647697099067e-08,
+      "loss": 0.1377,
+      "step": 4215
+    },
+    {
+      "epoch": 0.9498972033683499,
+      "grad_norm": 0.45698200368309266,
+      "learning_rate": 3.412902519217137e-08,
+      "loss": 0.1364,
+      "step": 4216
+    },
+    {
+      "epoch": 0.9501225110541583,
+      "grad_norm": 0.47121234535073564,
+      "learning_rate": 3.382294261959157e-08,
+      "loss": 0.1397,
+      "step": 4217
+    },
+    {
+      "epoch": 0.9503478187399668,
+      "grad_norm": 0.50536781568685,
+      "learning_rate": 3.351822942320754e-08,
+      "loss": 0.1491,
+      "step": 4218
+    },
+    {
+      "epoch": 0.9505731264257752,
+      "grad_norm": 0.5033394410768004,
+      "learning_rate": 3.3214885772215046e-08,
+      "loss": 0.1483,
+      "step": 4219
+    },
+    {
+      "epoch": 0.9507984341115836,
+      "grad_norm": 0.460921227692927,
+      "learning_rate": 3.2912911835049634e-08,
+      "loss": 0.1339,
+      "step": 4220
+    },
+    {
+      "epoch": 0.951023741797392,
+      "grad_norm": 0.48383243993925323,
+      "learning_rate": 3.261230777938607e-08,
+      "loss": 0.1532,
+      "step": 4221
+    },
+    {
+      "epoch": 0.9512490494832005,
+      "grad_norm": 0.4984808680944678,
+      "learning_rate": 3.231307377213833e-08,
+      "loss": 0.1562,
+      "step": 4222
+    },
+    {
+      "epoch": 0.951474357169009,
+      "grad_norm": 0.5030971544427753,
+      "learning_rate": 3.201520997946045e-08,
+      "loss": 0.1528,
+      "step": 4223
+    },
+    {
+      "epoch": 0.9516996648548174,
+      "grad_norm": 0.4521382543613778,
+      "learning_rate": 3.171871656674458e-08,
+      "loss": 0.1294,
+      "step": 4224
+    },
+    {
+      "epoch": 0.9519249725406258,
+      "grad_norm": 0.4621539139022736,
+      "learning_rate": 3.142359369862291e-08,
+      "loss": 0.1353,
+      "step": 4225
+    },
+    {
+      "epoch": 0.9521502802264342,
+      "grad_norm": 0.4968241853667539,
+      "learning_rate": 3.112984153896603e-08,
+      "loss": 0.1627,
+      "step": 4226
+    },
+    {
+      "epoch": 0.9523755879122426,
+      "grad_norm": 0.4835846870993006,
+      "learning_rate": 3.0837460250883186e-08,
+      "loss": 0.1453,
+      "step": 4227
+    },
+    {
+      "epoch": 0.9526008955980511,
+      "grad_norm": 0.49806221123746486,
+      "learning_rate": 3.0546449996723404e-08,
+      "loss": 0.1445,
+      "step": 4228
+    },
+    {
+      "epoch": 0.9528262032838595,
+      "grad_norm": 0.4622365541956114,
+      "learning_rate": 3.0256810938073534e-08,
+      "loss": 0.1336,
+      "step": 4229
+    },
+    {
+      "epoch": 0.953051510969668,
+      "grad_norm": 0.5083718135711455,
+      "learning_rate": 2.996854323575937e-08,
+      "loss": 0.1504,
+      "step": 4230
+    },
+    {
+      "epoch": 0.9532768186554764,
+      "grad_norm": 0.4696426708038953,
+      "learning_rate": 2.968164704984483e-08,
+      "loss": 0.1398,
+      "step": 4231
+    },
+    {
+      "epoch": 0.9535021263412848,
+      "grad_norm": 0.4957647803669633,
+      "learning_rate": 2.939612253963331e-08,
+      "loss": 0.1485,
+      "step": 4232
+    },
+    {
+      "epoch": 0.9537274340270933,
+      "grad_norm": 0.4976801047971032,
+      "learning_rate": 2.911196986366577e-08,
+      "loss": 0.1577,
+      "step": 4233
+    },
+    {
+      "epoch": 0.9539527417129017,
+      "grad_norm": 0.5047313617268352,
+      "learning_rate": 2.8829189179721552e-08,
+      "loss": 0.1448,
+      "step": 4234
+    },
+    {
+      "epoch": 0.9541780493987101,
+      "grad_norm": 0.48704566793067555,
+      "learning_rate": 2.8547780644818113e-08,
+      "loss": 0.1391,
+      "step": 4235
+    },
+    {
+      "epoch": 0.9544033570845185,
+      "grad_norm": 0.45381551822981764,
+      "learning_rate": 2.8267744415211296e-08,
+      "loss": 0.1295,
+      "step": 4236
+    },
+    {
+      "epoch": 0.9546286647703269,
+      "grad_norm": 0.5171129482131813,
+      "learning_rate": 2.7989080646394217e-08,
+      "loss": 0.1678,
+      "step": 4237
+    },
+    {
+      "epoch": 0.9548539724561355,
+      "grad_norm": 0.46855959342887893,
+      "learning_rate": 2.7711789493099495e-08,
+      "loss": 0.1329,
+      "step": 4238
+    },
+    {
+      "epoch": 0.9550792801419439,
+      "grad_norm": 0.4550251335595192,
+      "learning_rate": 2.743587110929563e-08,
+      "loss": 0.1311,
+      "step": 4239
+    },
+    {
+      "epoch": 0.9553045878277523,
+      "grad_norm": 0.4649516123124458,
+      "learning_rate": 2.716132564819035e-08,
+      "loss": 0.1472,
+      "step": 4240
+    },
+    {
+      "epoch": 0.9555298955135607,
+      "grad_norm": 0.47923186532440143,
+      "learning_rate": 2.688815326222838e-08,
+      "loss": 0.1415,
+      "step": 4241
+    },
+    {
+      "epoch": 0.9557552031993691,
+      "grad_norm": 0.4943590216368037,
+      "learning_rate": 2.661635410309199e-08,
+      "loss": 0.1446,
+      "step": 4242
+    },
+    {
+      "epoch": 0.9559805108851775,
+      "grad_norm": 0.4936773150917739,
+      "learning_rate": 2.6345928321701575e-08,
+      "loss": 0.1444,
+      "step": 4243
+    },
+    {
+      "epoch": 0.956205818570986,
+      "grad_norm": 0.44895547905801686,
+      "learning_rate": 2.6076876068213965e-08,
+      "loss": 0.1285,
+      "step": 4244
+    },
+    {
+      "epoch": 0.9564311262567944,
+      "grad_norm": 0.4804177263012208,
+      "learning_rate": 2.5809197492024372e-08,
+      "loss": 0.1497,
+      "step": 4245
+    },
+    {
+      "epoch": 0.9566564339426029,
+      "grad_norm": 0.4804905277345808,
+      "learning_rate": 2.554289274176419e-08,
+      "loss": 0.14,
+      "step": 4246
+    },
+    {
+      "epoch": 0.9568817416284113,
+      "grad_norm": 0.48656611905634656,
+      "learning_rate": 2.5277961965302633e-08,
+      "loss": 0.1465,
+      "step": 4247
+    },
+    {
+      "epoch": 0.9571070493142197,
+      "grad_norm": 0.45690808190376064,
+      "learning_rate": 2.5014405309746193e-08,
+      "loss": 0.1316,
+      "step": 4248
+    },
+    {
+      "epoch": 0.9573323570000282,
+      "grad_norm": 0.48327173788939704,
+      "learning_rate": 2.4752222921437807e-08,
+      "loss": 0.138,
+      "step": 4249
+    },
+    {
+      "epoch": 0.9575576646858366,
+      "grad_norm": 0.5068194329354152,
+      "learning_rate": 2.449141494595797e-08,
+      "loss": 0.1532,
+      "step": 4250
+    },
+    {
+      "epoch": 0.957782972371645,
+      "grad_norm": 0.46654742066027965,
+      "learning_rate": 2.423198152812306e-08,
+      "loss": 0.1356,
+      "step": 4251
+    },
+    {
+      "epoch": 0.9580082800574534,
+      "grad_norm": 0.483696186611939,
+      "learning_rate": 2.3973922811987295e-08,
+      "loss": 0.1472,
+      "step": 4252
+    },
+    {
+      "epoch": 0.9582335877432618,
+      "grad_norm": 0.47201416927326817,
+      "learning_rate": 2.3717238940840493e-08,
+      "loss": 0.1361,
+      "step": 4253
+    },
+    {
+      "epoch": 0.9584588954290704,
+      "grad_norm": 0.45781860941730246,
+      "learning_rate": 2.3461930057210037e-08,
+      "loss": 0.1371,
+      "step": 4254
+    },
+    {
+      "epoch": 0.9586842031148788,
+      "grad_norm": 0.4491060009907602,
+      "learning_rate": 2.320799630285947e-08,
+      "loss": 0.1297,
+      "step": 4255
+    },
+    {
+      "epoch": 0.9589095108006872,
+      "grad_norm": 0.48731042626093923,
+      "learning_rate": 2.2955437818788508e-08,
+      "loss": 0.1549,
+      "step": 4256
+    },
+    {
+      "epoch": 0.9591348184864956,
+      "grad_norm": 0.4646098385451307,
+      "learning_rate": 2.2704254745233577e-08,
+      "loss": 0.136,
+      "step": 4257
+    },
+    {
+      "epoch": 0.959360126172304,
+      "grad_norm": 0.4759653882078851,
+      "learning_rate": 2.2454447221667563e-08,
+      "loss": 0.1445,
+      "step": 4258
+    },
+    {
+      "epoch": 0.9595854338581125,
+      "grad_norm": 0.48335670581672063,
+      "learning_rate": 2.2206015386798673e-08,
+      "loss": 0.1395,
+      "step": 4259
+    },
+    {
+      "epoch": 0.9598107415439209,
+      "grad_norm": 0.47111700747094354,
+      "learning_rate": 2.1958959378572398e-08,
+      "loss": 0.1335,
+      "step": 4260
+    },
+    {
+      "epoch": 0.9600360492297294,
+      "grad_norm": 0.468816202711293,
+      "learning_rate": 2.1713279334169278e-08,
+      "loss": 0.1448,
+      "step": 4261
+    },
+    {
+      "epoch": 0.9602613569155378,
+      "grad_norm": 0.44960534206662445,
+      "learning_rate": 2.1468975390006587e-08,
+      "loss": 0.1305,
+      "step": 4262
+    },
+    {
+      "epoch": 0.9604866646013462,
+      "grad_norm": 0.45547617142439345,
+      "learning_rate": 2.1226047681737193e-08,
+      "loss": 0.1349,
+      "step": 4263
+    },
+    {
+      "epoch": 0.9607119722871547,
+      "grad_norm": 0.48149633710593354,
+      "learning_rate": 2.0984496344249596e-08,
+      "loss": 0.142,
+      "step": 4264
+    },
+    {
+      "epoch": 0.9609372799729631,
+      "grad_norm": 0.4636145891113348,
+      "learning_rate": 2.074432151166844e-08,
+      "loss": 0.1314,
+      "step": 4265
+    },
+    {
+      "epoch": 0.9611625876587715,
+      "grad_norm": 0.5067529814876942,
+      "learning_rate": 2.0505523317353727e-08,
+      "loss": 0.1564,
+      "step": 4266
+    },
+    {
+      "epoch": 0.9613878953445799,
+      "grad_norm": 0.4662581473086844,
+      "learning_rate": 2.0268101893901327e-08,
+      "loss": 0.1368,
+      "step": 4267
+    },
+    {
+      "epoch": 0.9616132030303883,
+      "grad_norm": 0.4766281613039792,
+      "learning_rate": 2.0032057373142453e-08,
+      "loss": 0.148,
+      "step": 4268
+    },
+    {
+      "epoch": 0.9618385107161969,
+      "grad_norm": 0.4715925046426043,
+      "learning_rate": 1.9797389886143658e-08,
+      "loss": 0.1492,
+      "step": 4269
+    },
+    {
+      "epoch": 0.9620638184020053,
+      "grad_norm": 0.44744281444466083,
+      "learning_rate": 1.956409956320737e-08,
+      "loss": 0.1383,
+      "step": 4270
+    },
+    {
+      "epoch": 0.9622891260878137,
+      "grad_norm": 0.49040603740334754,
+      "learning_rate": 1.933218653387081e-08,
+      "loss": 0.1517,
+      "step": 4271
+    },
+    {
+      "epoch": 0.9625144337736221,
+      "grad_norm": 0.4781738093055147,
+      "learning_rate": 1.91016509269068e-08,
+      "loss": 0.1477,
+      "step": 4272
+    },
+    {
+      "epoch": 0.9627397414594305,
+      "grad_norm": 0.5043088508617445,
+      "learning_rate": 1.8872492870322945e-08,
+      "loss": 0.153,
+      "step": 4273
+    },
+    {
+      "epoch": 0.962965049145239,
+      "grad_norm": 0.45938917413578406,
+      "learning_rate": 1.864471249136218e-08,
+      "loss": 0.1386,
+      "step": 4274
+    },
+    {
+      "epoch": 0.9631903568310474,
+      "grad_norm": 0.48259651582192653,
+      "learning_rate": 1.8418309916502787e-08,
+      "loss": 0.1432,
+      "step": 4275
+    },
+    {
+      "epoch": 0.9634156645168558,
+      "grad_norm": 0.4598535106472366,
+      "learning_rate": 1.819328527145725e-08,
+      "loss": 0.1241,
+      "step": 4276
+    },
+    {
+      "epoch": 0.9636409722026643,
+      "grad_norm": 0.46091011665324716,
+      "learning_rate": 1.7969638681173684e-08,
+      "loss": 0.1367,
+      "step": 4277
+    },
+    {
+      "epoch": 0.9638662798884727,
+      "grad_norm": 0.48351619131549517,
+      "learning_rate": 1.774737026983414e-08,
+      "loss": 0.1379,
+      "step": 4278
+    },
+    {
+      "epoch": 0.9640915875742812,
+      "grad_norm": 0.481262920354411,
+      "learning_rate": 1.752648016085684e-08,
+      "loss": 0.1506,
+      "step": 4279
+    },
+    {
+      "epoch": 0.9643168952600896,
+      "grad_norm": 0.4805939380195428,
+      "learning_rate": 1.7306968476893393e-08,
+      "loss": 0.1448,
+      "step": 4280
+    },
+    {
+      "epoch": 0.964542202945898,
+      "grad_norm": 0.5055025504837728,
+      "learning_rate": 1.708883533983019e-08,
+      "loss": 0.1463,
+      "step": 4281
+    },
+    {
+      "epoch": 0.9647675106317064,
+      "grad_norm": 0.49571269367735965,
+      "learning_rate": 1.6872080870788955e-08,
+      "loss": 0.1473,
+      "step": 4282
+    },
+    {
+      "epoch": 0.9649928183175148,
+      "grad_norm": 0.4533431039913623,
+      "learning_rate": 1.6656705190125078e-08,
+      "loss": 0.1274,
+      "step": 4283
+    },
+    {
+      "epoch": 0.9652181260033232,
+      "grad_norm": 0.46643306934522594,
+      "learning_rate": 1.6442708417428732e-08,
+      "loss": 0.1391,
+      "step": 4284
+    },
+    {
+      "epoch": 0.9654434336891318,
+      "grad_norm": 0.4655292603080353,
+      "learning_rate": 1.6230090671524312e-08,
+      "loss": 0.1403,
+      "step": 4285
+    },
+    {
+      "epoch": 0.9656687413749402,
+      "grad_norm": 0.505200988979893,
+      "learning_rate": 1.6018852070470437e-08,
+      "loss": 0.1443,
+      "step": 4286
+    },
+    {
+      "epoch": 0.9658940490607486,
+      "grad_norm": 0.49078232446343767,
+      "learning_rate": 1.5808992731560225e-08,
+      "loss": 0.1451,
+      "step": 4287
+    },
+    {
+      "epoch": 0.966119356746557,
+      "grad_norm": 0.4828868989290041,
+      "learning_rate": 1.5600512771320462e-08,
+      "loss": 0.1357,
+      "step": 4288
+    },
+    {
+      "epoch": 0.9663446644323654,
+      "grad_norm": 0.49321915716373294,
+      "learning_rate": 1.5393412305512446e-08,
+      "loss": 0.1406,
+      "step": 4289
+    },
+    {
+      "epoch": 0.9665699721181739,
+      "grad_norm": 0.4797986763655872,
+      "learning_rate": 1.518769144913168e-08,
+      "loss": 0.1368,
+      "step": 4290
+    },
+    {
+      "epoch": 0.9667952798039823,
+      "grad_norm": 0.4583811913091939,
+      "learning_rate": 1.4983350316406797e-08,
+      "loss": 0.1345,
+      "step": 4291
+    },
+    {
+      "epoch": 0.9670205874897907,
+      "grad_norm": 0.4615067525946092,
+      "learning_rate": 1.4780389020800923e-08,
+      "loss": 0.1322,
+      "step": 4292
+    },
+    {
+      "epoch": 0.9672458951755992,
+      "grad_norm": 0.4497627493649979,
+      "learning_rate": 1.4578807675011131e-08,
+      "loss": 0.1334,
+      "step": 4293
+    },
+    {
+      "epoch": 0.9674712028614076,
+      "grad_norm": 0.4835148026172114,
+      "learning_rate": 1.4378606390967609e-08,
+      "loss": 0.1441,
+      "step": 4294
+    },
+    {
+      "epoch": 0.9676965105472161,
+      "grad_norm": 0.481902922708207,
+      "learning_rate": 1.4179785279835045e-08,
+      "loss": 0.1431,
+      "step": 4295
+    },
+    {
+      "epoch": 0.9679218182330245,
+      "grad_norm": 0.4582081824677256,
+      "learning_rate": 1.3982344452011242e-08,
+      "loss": 0.1348,
+      "step": 4296
+    },
+    {
+      "epoch": 0.9681471259188329,
+      "grad_norm": 0.47063804644729257,
+      "learning_rate": 1.3786284017127949e-08,
+      "loss": 0.1429,
+      "step": 4297
+    },
+    {
+      "epoch": 0.9683724336046413,
+      "grad_norm": 0.474544321572788,
+      "learning_rate": 1.3591604084049747e-08,
+      "loss": 0.1406,
+      "step": 4298
+    },
+    {
+      "epoch": 0.9685977412904497,
+      "grad_norm": 0.47956333804497253,
+      "learning_rate": 1.3398304760875725e-08,
+      "loss": 0.144,
+      "step": 4299
+    },
+    {
+      "epoch": 0.9688230489762581,
+      "grad_norm": 0.47952835497451235,
+      "learning_rate": 1.3206386154937245e-08,
+      "loss": 0.1489,
+      "step": 4300
+    },
+    {
+      "epoch": 0.9690483566620667,
+      "grad_norm": 0.5011216096374138,
+      "learning_rate": 1.30158483727999e-08,
+      "loss": 0.1519,
+      "step": 4301
+    },
+    {
+      "epoch": 0.9692736643478751,
+      "grad_norm": 0.4619877849945653,
+      "learning_rate": 1.2826691520262114e-08,
+      "loss": 0.1335,
+      "step": 4302
+    },
+    {
+      "epoch": 0.9694989720336835,
+      "grad_norm": 0.4510567616253443,
+      "learning_rate": 1.2638915702355702e-08,
+      "loss": 0.1271,
+      "step": 4303
+    },
+    {
+      "epoch": 0.9697242797194919,
+      "grad_norm": 0.46742284679784957,
+      "learning_rate": 1.2452521023345598e-08,
+      "loss": 0.1353,
+      "step": 4304
+    },
+    {
+      "epoch": 0.9699495874053004,
+      "grad_norm": 0.4744154062862906,
+      "learning_rate": 1.2267507586729566e-08,
+      "loss": 0.1448,
+      "step": 4305
+    },
+    {
+      "epoch": 0.9701748950911088,
+      "grad_norm": 0.4365727581849454,
+      "learning_rate": 1.2083875495238761e-08,
+      "loss": 0.1246,
+      "step": 4306
+    },
+    {
+      "epoch": 0.9704002027769172,
+      "grad_norm": 0.46537369554131985,
+      "learning_rate": 1.1901624850837734e-08,
+      "loss": 0.1428,
+      "step": 4307
+    },
+    {
+      "epoch": 0.9706255104627256,
+      "grad_norm": 0.49077869002621655,
+      "learning_rate": 1.1720755754722757e-08,
+      "loss": 0.1494,
+      "step": 4308
+    },
+    {
+      "epoch": 0.9708508181485341,
+      "grad_norm": 0.4976622519358272,
+      "learning_rate": 1.1541268307324049e-08,
+      "loss": 0.1462,
+      "step": 4309
+    },
+    {
+      "epoch": 0.9710761258343426,
+      "grad_norm": 0.4936796887656191,
+      "learning_rate": 1.1363162608304112e-08,
+      "loss": 0.1477,
+      "step": 4310
+    },
+    {
+      "epoch": 0.971301433520151,
+      "grad_norm": 0.497374029101745,
+      "learning_rate": 1.1186438756558838e-08,
+      "loss": 0.1482,
+      "step": 4311
+    },
+    {
+      "epoch": 0.9715267412059594,
+      "grad_norm": 0.4698021005115719,
+      "learning_rate": 1.1011096850215842e-08,
+      "loss": 0.1455,
+      "step": 4312
+    },
+    {
+      "epoch": 0.9717520488917678,
+      "grad_norm": 0.4792243933647282,
+      "learning_rate": 1.083713698663641e-08,
+      "loss": 0.1389,
+      "step": 4313
+    },
+    {
+      "epoch": 0.9719773565775762,
+      "grad_norm": 0.4626205082490106,
+      "learning_rate": 1.0664559262413831e-08,
+      "loss": 0.1281,
+      "step": 4314
+    },
+    {
+      "epoch": 0.9722026642633846,
+      "grad_norm": 0.4874428688965878,
+      "learning_rate": 1.0493363773373677e-08,
+      "loss": 0.1551,
+      "step": 4315
+    },
+    {
+      "epoch": 0.9724279719491932,
+      "grad_norm": 0.5004355080951479,
+      "learning_rate": 1.0323550614574907e-08,
+      "loss": 0.1517,
+      "step": 4316
+    },
+    {
+      "epoch": 0.9726532796350016,
+      "grad_norm": 0.47262833063941007,
+      "learning_rate": 1.0155119880308483e-08,
+      "loss": 0.1352,
+      "step": 4317
+    },
+    {
+      "epoch": 0.97287858732081,
+      "grad_norm": 0.46940376415483154,
+      "learning_rate": 9.988071664097376e-09,
+      "loss": 0.1387,
+      "step": 4318
+    },
+    {
+      "epoch": 0.9731038950066184,
+      "grad_norm": 0.47903980564159626,
+      "learning_rate": 9.822406058697665e-09,
+      "loss": 0.146,
+      "step": 4319
+    },
+    {
+      "epoch": 0.9733292026924268,
+      "grad_norm": 0.4949962235080336,
+      "learning_rate": 9.658123156096599e-09,
+      "loss": 0.1465,
+      "step": 4320
+    },
+    {
+      "epoch": 0.9735545103782353,
+      "grad_norm": 0.4744343780732498,
+      "learning_rate": 9.4952230475151e-09,
+      "loss": 0.1436,
+      "step": 4321
+    },
+    {
+      "epoch": 0.9737798180640437,
+      "grad_norm": 0.45313837725329525,
+      "learning_rate": 9.333705823404981e-09,
+      "loss": 0.1364,
+      "step": 4322
+    },
+    {
+      "epoch": 0.9740051257498521,
+      "grad_norm": 0.48156914502453946,
+      "learning_rate": 9.17357157345089e-09,
+      "loss": 0.1423,
+      "step": 4323
+    },
+    {
+      "epoch": 0.9742304334356606,
+      "grad_norm": 0.493321796410043,
+      "learning_rate": 9.014820386569756e-09,
+      "loss": 0.1518,
+      "step": 4324
+    },
+    {
+      "epoch": 0.974455741121469,
+      "grad_norm": 0.47247014797296477,
+      "learning_rate": 8.85745235090968e-09,
+      "loss": 0.1374,
+      "step": 4325
+    },
+    {
+      "epoch": 0.9746810488072775,
+      "grad_norm": 0.4778581083662472,
+      "learning_rate": 8.701467553851317e-09,
+      "loss": 0.1459,
+      "step": 4326
+    },
+    {
+      "epoch": 0.9749063564930859,
+      "grad_norm": 0.4666557114127274,
+      "learning_rate": 8.54686608200761e-09,
+      "loss": 0.1432,
+      "step": 4327
+    },
+    {
+      "epoch": 0.9751316641788943,
+      "grad_norm": 0.46927749290389636,
+      "learning_rate": 8.393648021222666e-09,
+      "loss": 0.1287,
+      "step": 4328
+    },
+    {
+      "epoch": 0.9753569718647027,
+      "grad_norm": 0.4788321168845856,
+      "learning_rate": 8.241813456573156e-09,
+      "loss": 0.1285,
+      "step": 4329
+    },
+    {
+      "epoch": 0.9755822795505111,
+      "grad_norm": 0.483308961427759,
+      "learning_rate": 8.09136247236636e-09,
+      "loss": 0.143,
+      "step": 4330
+    },
+    {
+      "epoch": 0.9758075872363196,
+      "grad_norm": 0.503398696917994,
+      "learning_rate": 7.942295152142954e-09,
+      "loss": 0.1576,
+      "step": 4331
+    },
+    {
+      "epoch": 0.9760328949221281,
+      "grad_norm": 0.5132744653021303,
+      "learning_rate": 7.79461157867395e-09,
+      "loss": 0.1652,
+      "step": 4332
+    },
+    {
+      "epoch": 0.9762582026079365,
+      "grad_norm": 0.46056290280919454,
+      "learning_rate": 7.64831183396264e-09,
+      "loss": 0.1321,
+      "step": 4333
+    },
+    {
+      "epoch": 0.9764835102937449,
+      "grad_norm": 0.45311995911352104,
+      "learning_rate": 7.503395999244045e-09,
+      "loss": 0.1267,
+      "step": 4334
+    },
+    {
+      "epoch": 0.9767088179795533,
+      "grad_norm": 0.45587215436860634,
+      "learning_rate": 7.359864154984353e-09,
+      "loss": 0.1313,
+      "step": 4335
+    },
+    {
+      "epoch": 0.9769341256653618,
+      "grad_norm": 0.48113082998224244,
+      "learning_rate": 7.217716380881479e-09,
+      "loss": 0.1443,
+      "step": 4336
+    },
+    {
+      "epoch": 0.9771594333511702,
+      "grad_norm": 0.47691226845499113,
+      "learning_rate": 7.076952755864508e-09,
+      "loss": 0.1447,
+      "step": 4337
+    },
+    {
+      "epoch": 0.9773847410369786,
+      "grad_norm": 0.48006758958441803,
+      "learning_rate": 6.937573358094529e-09,
+      "loss": 0.1355,
+      "step": 4338
+    },
+    {
+      "epoch": 0.977610048722787,
+      "grad_norm": 0.47284929946531534,
+      "learning_rate": 6.799578264963802e-09,
+      "loss": 0.1344,
+      "step": 4339
+    },
+    {
+      "epoch": 0.9778353564085955,
+      "grad_norm": 0.47254716605333813,
+      "learning_rate": 6.662967553095756e-09,
+      "loss": 0.1451,
+      "step": 4340
+    },
+    {
+      "epoch": 0.978060664094404,
+      "grad_norm": 0.44905567061828916,
+      "learning_rate": 6.527741298345269e-09,
+      "loss": 0.1313,
+      "step": 4341
+    },
+    {
+      "epoch": 0.9782859717802124,
+      "grad_norm": 0.4691229636171862,
+      "learning_rate": 6.3938995757981125e-09,
+      "loss": 0.1367,
+      "step": 4342
+    },
+    {
+      "epoch": 0.9785112794660208,
+      "grad_norm": 0.45663032287457944,
+      "learning_rate": 6.2614424597720605e-09,
+      "loss": 0.1304,
+      "step": 4343
+    },
+    {
+      "epoch": 0.9787365871518292,
+      "grad_norm": 0.44357322752828715,
+      "learning_rate": 6.1303700238152245e-09,
+      "loss": 0.1224,
+      "step": 4344
+    },
+    {
+      "epoch": 0.9789618948376376,
+      "grad_norm": 0.48531056325011684,
+      "learning_rate": 6.00068234070772e-09,
+      "loss": 0.1401,
+      "step": 4345
+    },
+    {
+      "epoch": 0.979187202523446,
+      "grad_norm": 0.48866064423236033,
+      "learning_rate": 5.8723794824597226e-09,
+      "loss": 0.15,
+      "step": 4346
+    },
+    {
+      "epoch": 0.9794125102092545,
+      "grad_norm": 0.47173835292473587,
+      "learning_rate": 5.745461520313411e-09,
+      "loss": 0.1458,
+      "step": 4347
+    },
+    {
+      "epoch": 0.979637817895063,
+      "grad_norm": 0.4606684373862621,
+      "learning_rate": 5.6199285247415805e-09,
+      "loss": 0.1356,
+      "step": 4348
+    },
+    {
+      "epoch": 0.9798631255808714,
+      "grad_norm": 0.45008526887749,
+      "learning_rate": 5.495780565447917e-09,
+      "loss": 0.1312,
+      "step": 4349
+    },
+    {
+      "epoch": 0.9800884332666798,
+      "grad_norm": 0.4894920967180572,
+      "learning_rate": 5.373017711367001e-09,
+      "loss": 0.1485,
+      "step": 4350
+    },
+    {
+      "epoch": 0.9803137409524882,
+      "grad_norm": 0.5039444257173281,
+      "learning_rate": 5.2516400306648615e-09,
+      "loss": 0.1453,
+      "step": 4351
+    },
+    {
+      "epoch": 0.9805390486382967,
+      "grad_norm": 0.44672365989159696,
+      "learning_rate": 5.131647590737587e-09,
+      "loss": 0.1275,
+      "step": 4352
+    },
+    {
+      "epoch": 0.9807643563241051,
+      "grad_norm": 0.49591763264309763,
+      "learning_rate": 5.0130404582127144e-09,
+      "loss": 0.1374,
+      "step": 4353
+    },
+    {
+      "epoch": 0.9809896640099135,
+      "grad_norm": 0.4804636892409362,
+      "learning_rate": 4.895818698948396e-09,
+      "loss": 0.1442,
+      "step": 4354
+    },
+    {
+      "epoch": 0.9812149716957219,
+      "grad_norm": 0.48197632973429744,
+      "learning_rate": 4.779982378033676e-09,
+      "loss": 0.1415,
+      "step": 4355
+    },
+    {
+      "epoch": 0.9814402793815304,
+      "grad_norm": 0.47806439019911473,
+      "learning_rate": 4.6655315597876615e-09,
+      "loss": 0.1421,
+      "step": 4356
+    },
+    {
+      "epoch": 0.9816655870673389,
+      "grad_norm": 0.46145387110204644,
+      "learning_rate": 4.552466307760905e-09,
+      "loss": 0.1314,
+      "step": 4357
+    },
+    {
+      "epoch": 0.9818908947531473,
+      "grad_norm": 0.4526873753515115,
+      "learning_rate": 4.440786684734577e-09,
+      "loss": 0.1362,
+      "step": 4358
+    },
+    {
+      "epoch": 0.9821162024389557,
+      "grad_norm": 0.48922876618552663,
+      "learning_rate": 4.330492752719628e-09,
+      "loss": 0.1502,
+      "step": 4359
+    },
+    {
+      "epoch": 0.9823415101247641,
+      "grad_norm": 0.4789288370287609,
+      "learning_rate": 4.221584572958737e-09,
+      "loss": 0.1405,
+      "step": 4360
+    },
+    {
+      "epoch": 0.9825668178105725,
+      "grad_norm": 0.4517484644833375,
+      "learning_rate": 4.114062205924085e-09,
+      "loss": 0.1268,
+      "step": 4361
+    },
+    {
+      "epoch": 0.982792125496381,
+      "grad_norm": 0.4830389432928069,
+      "learning_rate": 4.0079257113190275e-09,
+      "loss": 0.1473,
+      "step": 4362
+    },
+    {
+      "epoch": 0.9830174331821895,
+      "grad_norm": 0.4423608570433734,
+      "learning_rate": 3.903175148077531e-09,
+      "loss": 0.1255,
+      "step": 4363
+    },
+    {
+      "epoch": 0.9832427408679979,
+      "grad_norm": 0.4599706442868318,
+      "learning_rate": 3.799810574363072e-09,
+      "loss": 0.1326,
+      "step": 4364
+    },
+    {
+      "epoch": 0.9834680485538063,
+      "grad_norm": 0.48203972353551805,
+      "learning_rate": 3.697832047570571e-09,
+      "loss": 0.1549,
+      "step": 4365
+    },
+    {
+      "epoch": 0.9836933562396147,
+      "grad_norm": 0.4835541316928907,
+      "learning_rate": 3.597239624325011e-09,
+      "loss": 0.1433,
+      "step": 4366
+    },
+    {
+      "epoch": 0.9839186639254232,
+      "grad_norm": 0.4788202007537194,
+      "learning_rate": 3.4980333604811567e-09,
+      "loss": 0.1496,
+      "step": 4367
+    },
+    {
+      "epoch": 0.9841439716112316,
+      "grad_norm": 0.46619578510949317,
+      "learning_rate": 3.4002133111246673e-09,
+      "loss": 0.1432,
+      "step": 4368
+    },
+    {
+      "epoch": 0.98436927929704,
+      "grad_norm": 0.5110741441195408,
+      "learning_rate": 3.303779530571538e-09,
+      "loss": 0.1605,
+      "step": 4369
+    },
+    {
+      "epoch": 0.9845945869828484,
+      "grad_norm": 0.4621865944157658,
+      "learning_rate": 3.208732072368104e-09,
+      "loss": 0.1346,
+      "step": 4370
+    },
+    {
+      "epoch": 0.9848198946686569,
+      "grad_norm": 0.4796600192057587,
+      "learning_rate": 3.1150709892899256e-09,
+      "loss": 0.1453,
+      "step": 4371
+    },
+    {
+      "epoch": 0.9850452023544654,
+      "grad_norm": 0.4678812341995167,
+      "learning_rate": 3.022796333344291e-09,
+      "loss": 0.1431,
+      "step": 4372
+    },
+    {
+      "epoch": 0.9852705100402738,
+      "grad_norm": 0.4786835908663167,
+      "learning_rate": 2.9319081557674377e-09,
+      "loss": 0.1459,
+      "step": 4373
+    },
+    {
+      "epoch": 0.9854958177260822,
+      "grad_norm": 0.5007725644868222,
+      "learning_rate": 2.8424065070262186e-09,
+      "loss": 0.1513,
+      "step": 4374
+    },
+    {
+      "epoch": 0.9857211254118906,
+      "grad_norm": 0.49174743123100595,
+      "learning_rate": 2.754291436817824e-09,
+      "loss": 0.1506,
+      "step": 4375
+    },
+    {
+      "epoch": 0.985946433097699,
+      "grad_norm": 0.46095084583775564,
+      "learning_rate": 2.6675629940689508e-09,
+      "loss": 0.1397,
+      "step": 4376
+    },
+    {
+      "epoch": 0.9861717407835074,
+      "grad_norm": 0.4674294425537989,
+      "learning_rate": 2.582221226936632e-09,
+      "loss": 0.1338,
+      "step": 4377
+    },
+    {
+      "epoch": 0.9863970484693159,
+      "grad_norm": 0.4623302703852915,
+      "learning_rate": 2.4982661828085175e-09,
+      "loss": 0.1356,
+      "step": 4378
+    },
+    {
+      "epoch": 0.9866223561551244,
+      "grad_norm": 0.4617577655604814,
+      "learning_rate": 2.415697908300929e-09,
+      "loss": 0.1356,
+      "step": 4379
+    },
+    {
+      "epoch": 0.9868476638409328,
+      "grad_norm": 0.4634214429564338,
+      "learning_rate": 2.3345164492616367e-09,
+      "loss": 0.1413,
+      "step": 4380
+    },
+    {
+      "epoch": 0.9870729715267412,
+      "grad_norm": 0.4702899505718087,
+      "learning_rate": 2.2547218507673606e-09,
+      "loss": 0.1315,
+      "step": 4381
+    },
+    {
+      "epoch": 0.9872982792125496,
+      "grad_norm": 0.48808317706093335,
+      "learning_rate": 2.1763141571248813e-09,
+      "loss": 0.1529,
+      "step": 4382
+    },
+    {
+      "epoch": 0.9875235868983581,
+      "grad_norm": 0.48563405143106425,
+      "learning_rate": 2.0992934118715948e-09,
+      "loss": 0.1413,
+      "step": 4383
+    },
+    {
+      "epoch": 0.9877488945841665,
+      "grad_norm": 0.4991973947467297,
+      "learning_rate": 2.0236596577738466e-09,
+      "loss": 0.1462,
+      "step": 4384
+    },
+    {
+      "epoch": 0.9879742022699749,
+      "grad_norm": 0.4501683046388823,
+      "learning_rate": 1.9494129368280432e-09,
+      "loss": 0.1204,
+      "step": 4385
+    },
+    {
+      "epoch": 0.9881995099557833,
+      "grad_norm": 0.47746210450182514,
+      "learning_rate": 1.876553290261207e-09,
+      "loss": 0.136,
+      "step": 4386
+    },
+    {
+      "epoch": 0.9884248176415918,
+      "grad_norm": 0.4546943022120308,
+      "learning_rate": 1.8050807585293095e-09,
+      "loss": 0.1297,
+      "step": 4387
+    },
+    {
+      "epoch": 0.9886501253274003,
+      "grad_norm": 0.5517091587017808,
+      "learning_rate": 1.7349953813183828e-09,
+      "loss": 0.1484,
+      "step": 4388
+    },
+    {
+      "epoch": 0.9888754330132087,
+      "grad_norm": 0.45122543164788703,
+      "learning_rate": 1.6662971975439645e-09,
+      "loss": 0.1248,
+      "step": 4389
+    },
+    {
+      "epoch": 0.9891007406990171,
+      "grad_norm": 0.48450689131031716,
+      "learning_rate": 1.5989862453522075e-09,
+      "loss": 0.1436,
+      "step": 4390
+    },
+    {
+      "epoch": 0.9893260483848255,
+      "grad_norm": 0.4756699371386352,
+      "learning_rate": 1.5330625621176598e-09,
+      "loss": 0.1383,
+      "step": 4391
+    },
+    {
+      "epoch": 0.9895513560706339,
+      "grad_norm": 0.4694479754875359,
+      "learning_rate": 1.468526184445762e-09,
+      "loss": 0.1311,
+      "step": 4392
+    },
+    {
+      "epoch": 0.9897766637564424,
+      "grad_norm": 0.4978229930092217,
+      "learning_rate": 1.4053771481711832e-09,
+      "loss": 0.1471,
+      "step": 4393
+    },
+    {
+      "epoch": 0.9900019714422508,
+      "grad_norm": 0.4905457833843764,
+      "learning_rate": 1.343615488357819e-09,
+      "loss": 0.1488,
+      "step": 4394
+    },
+    {
+      "epoch": 0.9902272791280593,
+      "grad_norm": 0.4681366072187746,
+      "learning_rate": 1.2832412393001814e-09,
+      "loss": 0.1355,
+      "step": 4395
+    },
+    {
+      "epoch": 0.9904525868138677,
+      "grad_norm": 0.45234859070769357,
+      "learning_rate": 1.2242544345211772e-09,
+      "loss": 0.1287,
+      "step": 4396
+    },
+    {
+      "epoch": 0.9906778944996761,
+      "grad_norm": 0.46650840082669476,
+      "learning_rate": 1.1666551067746058e-09,
+      "loss": 0.1481,
+      "step": 4397
+    },
+    {
+      "epoch": 0.9909032021854846,
+      "grad_norm": 0.49602833209947034,
+      "learning_rate": 1.1104432880429394e-09,
+      "loss": 0.1405,
+      "step": 4398
+    },
+    {
+      "epoch": 0.991128509871293,
+      "grad_norm": 0.46169760847092306,
+      "learning_rate": 1.0556190095384333e-09,
+      "loss": 0.1366,
+      "step": 4399
+    },
+    {
+      "epoch": 0.9913538175571014,
+      "grad_norm": 0.47331739771221604,
+      "learning_rate": 1.0021823017028475e-09,
+      "loss": 0.1438,
+      "step": 4400
+    },
+    {
+      "epoch": 0.9915791252429098,
+      "grad_norm": 0.49149463979989205,
+      "learning_rate": 9.501331942080029e-10,
+      "loss": 0.1533,
+      "step": 4401
+    },
+    {
+      "epoch": 0.9918044329287182,
+      "grad_norm": 0.5040049366138424,
+      "learning_rate": 8.994717159546695e-10,
+      "loss": 0.1517,
+      "step": 4402
+    },
+    {
+      "epoch": 0.9920297406145268,
+      "grad_norm": 0.4588404544201829,
+      "learning_rate": 8.501978950734014e-10,
+      "loss": 0.1317,
+      "step": 4403
+    },
+    {
+      "epoch": 0.9922550483003352,
+      "grad_norm": 0.4605093760903773,
+      "learning_rate": 8.023117589237017e-10,
+      "loss": 0.1315,
+      "step": 4404
+    },
+    {
+      "epoch": 0.9924803559861436,
+      "grad_norm": 0.4867784316909424,
+      "learning_rate": 7.558133340954121e-10,
+      "loss": 0.1519,
+      "step": 4405
+    },
+    {
+      "epoch": 0.992705663671952,
+      "grad_norm": 0.49264063501342575,
+      "learning_rate": 7.10702646406769e-10,
+      "loss": 0.1369,
+      "step": 4406
+    },
+    {
+      "epoch": 0.9929309713577604,
+      "grad_norm": 0.48651414304036755,
+      "learning_rate": 6.669797209069018e-10,
+      "loss": 0.1472,
+      "step": 4407
+    },
+    {
+      "epoch": 0.9931562790435688,
+      "grad_norm": 0.45973115163132117,
+      "learning_rate": 6.246445818727798e-10,
+      "loss": 0.1294,
+      "step": 4408
+    },
+    {
+      "epoch": 0.9933815867293773,
+      "grad_norm": 0.46247424902217044,
+      "learning_rate": 5.836972528119878e-10,
+      "loss": 0.1445,
+      "step": 4409
+    },
+    {
+      "epoch": 0.9936068944151858,
+      "grad_norm": 0.46562589659047027,
+      "learning_rate": 5.44137756460783e-10,
+      "loss": 0.1295,
+      "step": 4410
+    },
+    {
+      "epoch": 0.9938322021009942,
+      "grad_norm": 0.45535410008693744,
+      "learning_rate": 5.059661147852057e-10,
+      "loss": 0.1388,
+      "step": 4411
+    },
+    {
+      "epoch": 0.9940575097868026,
+      "grad_norm": 0.47495739736859655,
+      "learning_rate": 4.691823489805236e-10,
+      "loss": 0.1413,
+      "step": 4412
+    },
+    {
+      "epoch": 0.994282817472611,
+      "grad_norm": 0.47225570583290694,
+      "learning_rate": 4.3378647947150965e-10,
+      "loss": 0.14,
+      "step": 4413
+    },
+    {
+      "epoch": 0.9945081251584195,
+      "grad_norm": 0.4726742313213338,
+      "learning_rate": 3.9977852591188694e-10,
+      "loss": 0.1416,
+      "step": 4414
+    },
+    {
+      "epoch": 0.9947334328442279,
+      "grad_norm": 0.4782536098463734,
+      "learning_rate": 3.671585071854389e-10,
+      "loss": 0.1392,
+      "step": 4415
+    },
+    {
+      "epoch": 0.9949587405300363,
+      "grad_norm": 0.4899719857969558,
+      "learning_rate": 3.3592644140434393e-10,
+      "loss": 0.1427,
+      "step": 4416
+    },
+    {
+      "epoch": 0.9951840482158447,
+      "grad_norm": 0.48595091219025266,
+      "learning_rate": 3.0608234591084083e-10,
+      "loss": 0.1429,
+      "step": 4417
+    },
+    {
+      "epoch": 0.9954093559016532,
+      "grad_norm": 0.48108261570009553,
+      "learning_rate": 2.776262372761185e-10,
+      "loss": 0.1504,
+      "step": 4418
+    },
+    {
+      "epoch": 0.9956346635874617,
+      "grad_norm": 0.4698281725163803,
+      "learning_rate": 2.505581313011485e-10,
+      "loss": 0.1382,
+      "step": 4419
+    },
+    {
+      "epoch": 0.9958599712732701,
+      "grad_norm": 0.4667519344743465,
+      "learning_rate": 2.2487804301557503e-10,
+      "loss": 0.1448,
+      "step": 4420
+    },
+    {
+      "epoch": 0.9960852789590785,
+      "grad_norm": 0.46820185366520156,
+      "learning_rate": 2.0058598667854755e-10,
+      "loss": 0.1415,
+      "step": 4421
+    },
+    {
+      "epoch": 0.9963105866448869,
+      "grad_norm": 0.4779993256669945,
+      "learning_rate": 1.776819757787207e-10,
+      "loss": 0.1353,
+      "step": 4422
+    },
+    {
+      "epoch": 0.9965358943306953,
+      "grad_norm": 0.4723393911423046,
+      "learning_rate": 1.561660230336992e-10,
+      "loss": 0.1444,
+      "step": 4423
+    },
+    {
+      "epoch": 0.9967612020165038,
+      "grad_norm": 0.4768298410372807,
+      "learning_rate": 1.3603814039031547e-10,
+      "loss": 0.1386,
+      "step": 4424
+    },
+    {
+      "epoch": 0.9969865097023122,
+      "grad_norm": 0.44014844744591247,
+      "learning_rate": 1.1729833902518473e-10,
+      "loss": 0.1202,
+      "step": 4425
+    },
+    {
+      "epoch": 0.9972118173881207,
+      "grad_norm": 0.4881840586497598,
+      "learning_rate": 9.994662934387223e-11,
+      "loss": 0.1442,
+      "step": 4426
+    },
+    {
+      "epoch": 0.9974371250739291,
+      "grad_norm": 0.4844812645696949,
+      "learning_rate": 8.398302098061583e-11,
+      "loss": 0.141,
+      "step": 4427
+    },
+    {
+      "epoch": 0.9976624327597375,
+      "grad_norm": 0.4619719293822893,
+      "learning_rate": 6.94075227999913e-11,
+      "loss": 0.139,
+      "step": 4428
+    },
+    {
+      "epoch": 0.997887740445546,
+      "grad_norm": 0.4570075187656965,
+      "learning_rate": 5.62201428946918e-11,
+      "loss": 0.1362,
+      "step": 4429
+    },
+    {
+      "epoch": 0.9981130481313544,
+      "grad_norm": 0.46402279082792575,
+      "learning_rate": 4.44208885877484e-11,
+      "loss": 0.1313,
+      "step": 4430
+    },
+    {
+      "epoch": 0.9983383558171628,
+      "grad_norm": 0.46837914976734196,
+      "learning_rate": 3.400976643030962e-11,
+      "loss": 0.1387,
+      "step": 4431
+    },
+    {
+      "epoch": 0.9985636635029712,
+      "grad_norm": 0.49390204409988,
+      "learning_rate": 2.498678220386186e-11,
+      "loss": 0.148,
+      "step": 4432
+    },
+    {
+      "epoch": 0.9987889711887796,
+      "grad_norm": 0.4884904018494601,
+      "learning_rate": 1.735194091800896e-11,
+      "loss": 0.1494,
+      "step": 4433
+    },
+    {
+      "epoch": 0.9990142788745882,
+      "grad_norm": 0.4828464055068492,
+      "learning_rate": 1.1105246812137538e-11,
+      "loss": 0.138,
+      "step": 4434
+    },
+    {
+      "epoch": 0.9992395865603966,
+      "grad_norm": 0.4629521120856078,
+      "learning_rate": 6.246703355139438e-12,
+      "loss": 0.1439,
+      "step": 4435
+    },
+    {
+      "epoch": 0.999464894246205,
+      "grad_norm": 0.45979903985531045,
+      "learning_rate": 2.7763132445790543e-12,
+      "loss": 0.1297,
+      "step": 4436
+    },
+    {
+      "epoch": 0.9996902019320134,
+      "grad_norm": 0.48535911253263453,
+      "learning_rate": 6.940784075259999e-13,
+      "loss": 0.1444,
+      "step": 4437
+    },
+    {
+      "epoch": 0.9999155096178218,
+      "grad_norm": 0.4531440552581044,
+      "learning_rate": 0.0,
+      "loss": 0.125,
+      "step": 4438
+    },
+    {
+      "epoch": 0.9999155096178218,
+      "step": 4438,
+      "total_flos": 995060424065024.0,
+      "train_loss": 0.16356557928304405,
+      "train_runtime": 22100.2501,
+      "train_samples_per_second": 12.853,
+      "train_steps_per_second": 0.201
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 4438,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 995060424065024.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}